Thursday, February 17, 2011

Fact-Checking the Fact Checker

"Why are all the speakers white? ... All of you are white. Go to hell."

By now, everyone has heard about John Wiley Price's racist outburst:

"Dallas County Commissioner John Wiley Price smiled, glared and rolled his eyes as he listened to several speakers Tuesday accuse him of being deceitful and duplicitous in his behind-the-scenes maneuvering to replace the county's popular elections administrator [Bruce Sherbet]. But when the final speaker, Dallas lawyer Jeff Turner, took the podium and referred to Price repeatedly as the county's 'chief mullah' and used the word 'tribal' in his remarks, Price's famous temper got the better of him."

This part of the story is easily dealt with. Was John Wiley Price out of line? Yes. Was he provoked? Yes. Does that justify his outburst? No.

After the jump, an important detail that was lost in the colorful exchange.

Richardson's own Bill McCalpin picks up the lost detail on his website

"Dallas County Commissioner John Wiley Price stated in a recent Dallas Morning News Q&A that the number of registered voters in Dallas County was larger in 1998 than in 2010."

Fact-checking this claim, McCalpin calls this "Literally true, yet quite misleading." McCalpin claims it's misleading because the number of registered voters fluctuates cyclically. It goes up when new voters register for Presidential elections. It goes down when the voter rolls get purged of voters who have moved out of Dallas County. Depending on the end points for comparison that you choose, you can show either increasing or decreasing voter registrations. Using "statistical inference," McCalpin himself judges "that the actual number of eligible voters in Dallas County probably increased by 30,000 to 50,000" between 1998 and 2010.

Let's assume that McCalpin's inference is correct. Does that mean John Wiley Price has no case? I'm not so sure. Dallas County has about 1.1 million registered voters out of population of about 2.4 million (about a 45% registration rate). Dallas County's population increased by about 300,000 between 1998 and 2010. If the county elections administrator just continued to register voters in the same ratio as before, you'd expect to see an increase of about 135,000 registered voters. McCalpin's "statistical inference" is that the actual number is more like "30,000 to 50,000." That's a pretty big gap, even if you grant that the state has gotten better at purging the voter rolls of ineligible voters and has gotten no better at getting new voters to register.

Maybe Bill McCalpin can explain away that gap. Maybe Bruce Sherbet can. John Wiley Price presumably believes the gap is real and blames the elections administrator for inadequate new voter registration. Or maybe John Wiley Price isn't really interested in explanations. Maybe he just wants to see a higher rate of voter registration and isn't happy with Sherbet's failure to deliver it.

Because McCalpin fails to take population growth into account, I have to judge the fact-checker's analysis of the numbers to also be, at best, "literally true, yet quite misleading."

By the way, on Monday the Richardson City Council presented a proclamation to Bruce Sherbet for his career achievements as Dallas County Elections Administrator. He was lauded for his fairness, impartiality, ethics, efficiency, dignity, respect and credibility. I hope John Wiley Price took notice. Without ever mentioning his name, without ever using a single four-letter word, the Richardson City Council used him to demonstrate how to diplomatically tell someone else to "Go to hell."

1 comment:

Mark Steger said...

In the above, I used estimates from the US Census Bureau. Now that the official numbers are out, it looks like the estimates were off. We should use the official numbers going forward (while remembering that the JWP outburst happened before the official numbers were released). Of course, for 1998, all we'll ever have are estimates. The Census Bureau had estimated that 28.2% of the population is under 18 years of age. I haven't seen what the official numbers are or whether the percentage has significantly gone up in ten years. Obviously, that could have some effect. I don't know if the Census Bureau estimates what percentage of the population are citizens eligible to vote. Obviously, that could have some effect. I haven't redone the math, but that estimated growth in population of 300,000 might be more like 150,000 in fact. That would suggest 67,500 more voters (using the same 45% from before), if the registration rate had stayed the same between 1998 and 2010. Closer to's statistical inference of 30,000 to 50,000 actual increase, but still larger. Fudge that 67,500 down a bit because of a greater increase in non-citizens and under-18, and maybe registration is keeping up with population growth. But 'keeping up' doesn't appear to be enough for JWP (many bosses won't accept just staying in place - they want progress). Even with generous estimates, it's hard to argue that Dallas County has made much, if any, progress in getting more eligible voters actually registered. That appears to be the root of JWP's complaint against Bruce Sherbet.