Lies, damn lies and statistics

091315marktwain

Mark Twain once said “There are three kinds of lies. There are lies, damn lies and statistics.”  With Labor Day passed, I see some candidates for local offices are misusing statistics to make claims that don’t quite match the facts.

Misuse of statistics, sadly, are emerging in the sheriff’s department race and come from a candidate who wants to be the chief law enforcement officer for Floyd County.

Former Bailiff and independent sheriff’s candidate Jimmy Howery posted this on Facebook recently:

In 2014 we had a roughly a 15% criminal clearance rate in Floyd County. This means that if a crime occurs in your house, there is an 85% chance an arrest won’t be made. Please think about the seriousness of that fact. This number is slightly up from a whopping 6.8% arrest rate in 2013.

No source was cited for the the claim. Accurate statistics on Virginia law enforcement comes from the Incident Based Reporting (IBR) central repository that the Virginia State Police has used to document criminal statistics as part of The Uniform Crime Reporting (UCR) system since 1994.

In its latest report issued for 2014, the State Police told Virginia’s General Assembly that “The IBR information has become the primary vehicle used to evaluate levels of criminal activity in jurisdictions throughout Virginia.”

Several years ago, the State Police found that “case clearance rates” cannot be accurately reported because the justice system often drops some cases as part of plea deals or in negotiations.  A person can be convicted on multiple charges but an agreement can make it appear that some cases were not cleared.

The 2014 report says:

This report no longer contains clearance rates. The IBR system does not accurately reflect the volume of clearances each department produces in a given year. Additionally, the number of arrests in a jurisdiction does not provide precise offense clearance information in that one person arrested could clear multiple offenses.

Clearance rates have not been reported for many of the past years because of the problems cited.

In addition, statistics compiled use a model that determines rates “per 100,000 residents.”  Floyd County’s population was 15,641 in 2014 and has never been close to 100,000.  So figures on its activities should be adjusted to reflect a smaller population and, if not so, can be misleading when compared to areas with higher populations.  Montgomery County’s population, for example, is twice as high.

In a normal statistical report that uses the “per 100,000” model, Floyd County had 1,547 criminal incidents in 2014 but when that is adjusted to the county’s actual population, IBR shows the county handled 242 incidents last year.  The State Police handled 65 in the county for the same period.

I cover the Circuit Court sessions of the county for The Floyd Press.  According to the claim posted on Facebook, “roughly 15 percent” of the cases in Floyd in 2014 would be just 36 cases.  I’ve seen that many local cases cleared in two or three months in court.

Records of the Circuit Court cases maintained by the Commonwealth Virginia show 187 cases resulting in convictions in Floyd County in 2014 and 116 this year.  The claim that the Floyd County Sheriff’s Department cleared only 6.8 percent of its cases in 2013 is not supported by court records which show more than 60 percent cleared that year.

Floyd County Circuit Court now is in session every Tuesday, more than double the schedule of just three years ago.  Prisoners from Floyd County are at an all-time high at the New River Valley Regional Jail and have topped more than 100 recently.  When the jail was established the estimates for Floyd were 25 or less.  Additional transportation is also required to handle the number of defendants sent to jail and prison from Floyd County Circuit Court.

The claim has also been made that only 10 DUI arrests were made by sheriff’s deputies in the last year.  An examination of court records shows more than 100 DUI arrests in Floyd County in the past year by both sheriff’s deputies and state troopers.  Many were arrest of drivers with multiple DUI’s.  I serve as Floyd County’s member on the Alcohol Safety Action Committee office in Christiansburg and have access to the records.

As Mark Twain said, statistics can be one of the three great lies.  When you see a statistic used in a political campaign and the source is not cited, question those numbers.  When you look at the number of cases investigated by the Floyd County Sheriff’s Department I see the number that appear to have been cleared is, yes, about 15 percent when applied to a “per 100,000” population but when adjusted for the actual population of the county, the clearance rate would be more than half of the incidents handled by the Sheriff’s Department.

We saw this kind of statistical gamesmanship used by Doug Weddle, the Floyd County resident and Montgomery County police officer who ran, and lost, to Brian Craig in the GOP primary earlier this year.  Weddle’s credibility took another hit after he told the public and reporters on primary election night that he endorsed and supported Craig, then went back on his word and is helping Howery.

Current Sheriff Shannon Zeman, who is retiring, fired Howery, a retired state police officer, as a part-time bailiff before the primary election after a letter from the now candidate appeared in The Floyd Press endorsed Weddle and made claims about operations of the department that were not true.  The endorsement of Weddle was not a factor in Howery’s dismissal.

As a reporter, I do not endorse or support any candidate but if unsubstantiated claims are made as “facts” made as claims I will investigate those claims.  If Howery will provide a source for his statistical claims, I will check that source to see how the figures were determined.  What is needed are names, dates and specific cases.  Using terms like “roughly” or “I understand that” or “some people have told me” is hearsay at best.

This looks to be a testy and nasty race for one of the most important elected positions in Floyd County.  In the end, the voters will make the choice.

(Edited to add additional information)

One Response to Lies, damn lies and statistics

  1. Doug, when you note ‘Records of the Circuit Court cases maintained by the Commonwealth Virginia show 187 cases resulting in convictions in Floyd County in 2014 and 116 this year’, do those convictions correspond to the Group A offenses? (You may want to specify that the figure 242 represents that Group.)