Misusing data to support personal agendas

Merritt Clifton is at it again -- and once again fudging numbers to promote his own personal agendas. 

In this month's issue of Animal People, Clifton's own newspaper where he is the sole writer and editor, Clifton releases his updated information on the state of shelter euthanasias in the U.S.

Not all of Clifton's information is garbage -- although enough of it is to even make me question the parts that seem accurate. He has certainly pulled ou this own conclusions in order to support his own personal agendsa

The article starts with a misleading headline "Decade of adoption focus fails to reduce shelter killing" -- even though in spite of pushes by proponents of the No Kill Movement, most city run shelters have not instituted much emphasis on adoptions. 

But this isn't the biggest problem with Clifton's data analysis.

Clifton is an advocate of breed bans. For reasons I don't know or understand he has continued to be maybe the loan torch bearer in the animal welfare community in support of breed bans.  And this document appears to be no exception.

In a little sidebar on the third page of the document he pulls out 11 cities (seemingly at random) and compares the euthanasia rates for 'pit bulls' in these shelters (3 of the cities have breed specific legislation, the other 8 do not).  The data also shows San Francisco's numbers before and after the city's ruling to mandate the spay/neuter all pit bulls.  Here is Clifton's analysis of the numbers:

"Philadelphia appears to kill relatively few pit bulls as a share of shelter killing chiefly because it has the highest total shelter killing on the chart: 19.9 per 1,000 people. Note that pit bulls are a much smaller part of the total killing in Cincinnati (which has a breed ban) than in either Milwaukee or Indianapolis, two cities with a very similar demographic profile."

The premise here appears to be if a city passes breed specific policies, then the number of pit bulls killed goes down, and that's a good thing. But once again, Clifton has manipulated data to support his point...not developed a point based on the actual data.

San Francisco

Let's start with San Francisco. According to Clifton's numbers, before instituting their MSN policy for pit bulls, San Francisco killed 452 pit bulls -- making up 43% of the total dogs killed in the shelter. Meanwhile, post the MSN, they killed 304 pit bulls, making up 29% of the dogs killed.

It looks like it's working great.

However, if you extrapolate from his data the total number of dogs killed, it tells a slightly different story. If the 304 pit bulls makes up 29% of the total kills, then San Francisco killed a total of 1,048 dogs post MSN.  If the 462 is 43% of the killed dogs, then a total of 1,051 dogs were killed a year prior to the MSN.

Essentially, the only thing that has changed is not a decrease in killing (as Clifton infers from the data), but a change in what type of dogs are killed. Whether this is a reflection of how dogs are classified at the shelter (were a lot more dogs classified as 'boxer mixes' vs 'pit bull' mixes in order to show "effectiveness at the shelter level) or did many of the irresponsible owners just change the type of dog they owned to avoid being harrassed by the law?  It seems Clifton doesn't seek to pursue this question - or even question it at all.  But it sure seems highly relevant.

Cincinnati, Milwaukee and Indianapolis

Meanwhile, Clifton uses a comparison of Cincinnati (a city with a breed ban) and both Milwaukee and Indianapolis as another example of how a city with a breed ban will kill fewer pit bulls than other cities.  But is this true?  From the chart, Clifton's assertion seems true as pit bulls make up a lower percentage of dogs killed than in the other two cities.

But wait.

Until as recently as this past March, Indianapolis' city shelter had a policy that kept them from adopting any pit bulls out of the city shelter. While this has changed in the past year under new shelter management, it certainly was in place with Clifton's current numbers. So an increase in percentage of pit bulls being killed at the shelter would be expected as an incredibly high percentage of the dogs would get killed in the shelter because POLICY prevented them from getting out alive. Clifton isn't a dumb man, and I'd be stunned if he didn't know this information. But he conveniently leaves this out of his analysis.

Meanwhile, when you look at Cincinnati compared to Milwuakee, you may notice something else significant. Even looking at the number Clifton posts, you see that even though pit bulls make up 18% of Milwaukee's euthanized dogs, the city is still killing 37% fewer pit bulls per thousand residents than the city of Cincinnati.  So what gives?

Well, again, if you extrapolate total numbers of dogs killed, you'll see that Milwaukee killed 7,988 total dogs during the timeframe available -- at a pace of 4.7 dogs per 1,000 people. Cincinnati on the other hand killed 11,210 dogs in that same time period -- a pace of 13.5 per 1,000 people*. So it appears that Cincinnati's "success" comes not from killing fewer 'pit bulls', but from the massive killing of dogs regardless of breed. This is also not something that would be a foreign concept to Clifton, who stated this exact scenerio as why the percentage of pit bulls killed in Philadelphia are so low - -and yet not only leaves this out of his "analysis" of the Cincinnati/Milwaukee comparison, he also somewhat conveniently leaves both cities out of the his overall shelter killing analysis that would make easy comparisons impossible.

* It should be noted here that I have no idea what Clifton is using for his population totals for these two cities.  While I can get there mathmatically by dividing the number of dogs killed by the # killed per 1,000 people, I don't know where he got these figures. Cincinnati's population of 836,000 appears to be the population of Hamilton County.  Milwaukee's population of 1.48 million is close to that of the entire Milwuakee MSA -- which they may provide service for.  For the sake of this whole post, I'm just using Clifton's own data.

Meanwhile, Denver -- who has had a breed ban since 1987 -- seems to have INCREDIBLE numbers based on Clifton's report -- killing only 275 pit bulls at a rate of .14 per 1,000 people and making up only 3% of the kill rate.  However, again, Denver's total numbers 9,167 killd dogs and a 4.6 kill rate are very similar to that of Milwaukee -- a city that doesn't have breed specific regulations.  It would only make sense that a city that has had a breed ban on the books for 20 years would not kill all that many dogs of that breed -- but it still hasn't seemed to improve the shelter's overall kill rate.   Meanwhile, it should be noted that Clifton's pit bull kill estimates seem extremely low compared to what most media reports give for the city and their pit bull kill numbers. These numbers for Denver are also conveniently left off of Clifton's overall report.

Why This matters

If we are really ever going to achieve a status of No Kill in our cities, we have to be HONEST with our numbers. We need to avoid the trickery and deceipt by people with agenda bias. We must be honest with the stats, and really look at what WORKS to create No Kill communities. No hidden agendas. No slight of hand statistics.  Just open, honest, dialogue.

This is why I think it's so incredible when cities provide open access to their information.  Go to cities provide open access to their information online for everyone to see -- like in Charlottesville, Va and Reno, NV. (Interestingly, Clifton doesn't include the numbers for Charlottesville even though the info is easily available).

When people like Clifton use the information to mislead the public, it hurts the animals. It hurts them by encouraging policy that is not helpful to the cause...and in fact, makes the situation worse. 

It's time for everyone to put their agendas aside and to push forward with policies that WORK to make communities more livable for both people and the animals. 

Comments

[Rinalia]

I really, really, really dislike that man. At the very least, us folks in the ar movement tend to find his publication laughable at best. He's an embarrassment.

[SocialMange]

Did Clifton invent any new breeds in this "study"? The batch of invented breeds he had in his last one gave me hours of entertainment.

[EmilyS]

and of course it is based on the whole assumption that a dog identified as a "pit bull" is actually an APBT/AST/SBT. Which as we know is complete b.s.

[PAMM - People Against Moronic Merritt]

What a f'in idiot...Merrit - can you and Kory Nelson, Tom Skelton and that dick from Canada just die already?

[Karen Delise]

Clifton not only doesn't know how to collect data, or analyze it - he is very deliberate in what data he collects (or ignores).
Many years ago, while writing my first book, Fatal Dog Attacks, I came across the old Clifton fatal dog attack report. I noticed he had two cases of fatal dog attacks that I did not have, but I had about 70 cases he did not have. I was very naive about all this at the time and called him up to ask him about the cases I couldn't find.
Clifton could not give me the information, (I can't remember the reason why). But, naive as I was, I asked him if he wanted the 70 cases I knew about that he was missing.
His answer- NO - NOT INTERESTED !!
That totally baffled me at the time -
But now I have a much better understanding of what consitutes a "raving nut case."
I can only speculate at the total disregard for accuracy and errors that can be found in his other "statistical" reports.
Karen Delise

[EmilyS]

Nelson, Skelton, Merritt, whatshername from dogbite.org, PBurns, all share a common trait. They get off on dog bites in a perverse, almost sexual way. Just read the way they write about this subject with that heightened excited hysteria: it's like some kind of pornography for them.

[MichelleD]

PAMM - That dick from Canada didn't die but he DID kill someone!?!??!!?

EmilyS - why did you have to put that imagine in my head!? ;-)

[selwyn marock]

PAMM's posting is undoubtably the best.

[email protected]

[selwyn marock]

Sorry EmilyS your posting is also excellant,just add Pizzano and Gallagher from Dade county,they get it on by cutting off the poor family pet's head.Sicko's

[email protected]

[Selma]

Gee, Brent, that stuff's difficult...

Seriously, Clifton wouldn't know real data (or how to use it) if it came up and bit him in the ass.

There's a good smackdown over at Lassie, Get Help in a few stellar posts. Put Clifton in the search box on the right sidebar near the top.

http://lassiegethelp.blogspot.com/

Surely even the lowest of the low-info types out there don't bother with him anymore.

[EmilyS]

oh Selma, I refer you to the personages I mention in my comment above, at least one of whom touts himself as honest and expert

[Jennifer]

My experience with Clifton was similar to Karen Delise's. Many years ago when I was first getting started with this sort of study, I was trying to understand his data collection method and innocently asked Clifton if he could provide me with his actual data. You know, some sort of list or chart or something that I could use to research and confirm the "bites" he'd summed up in that neat little chart.

First he told me that he couldn't find the information. But in that same email, he bragged about how he shared his data with so many authorities in the animal control/welfare/rights field. Hearing this, I suggested I could get the data from one of the people he'd already given it to, if he could suggest the name of someone.

Then he backtracked and said that he had the data, but that it was hard copy in a big file cabinet, and so it would be hard to dig out. I was confused, and asked how he could have so much trouble giving me the data, yet he had apparently shared it with others on previous occasions.

At this point, he informed me that as I was an insignificant nobody (not the words he used--he was much more rude), quite possibly with a hidden agenda to destroy his data, I could take a flying leap. I was floored by his hostile reaction to my inquiry, especially since he had adamantly insisted that his data had been collected scientifically and analyzed honestly.

My experience with real researchers, even as an insignificant nobody, is quite the opposite--assuming the matter is not confidential, they're happy to share their data (which is readily available) and their insights, and get feedback in order to refine their conclusions.

Needless to say, I consider Clifton's "reports" half fantasy, much like his self-aggrandizing attitude.

[KerryAnn May]

Below I've cut and pasted two email exchanges with Merritt. I basically asked him to clarify the claim that Terre Haute had killed 78 animals. This is what ensued. The part after >>>>> is what I wrote, and then he responds. Until then, I had never interacted with the man and was quite surprised to be treated with so rudely!

-----Original Message-----
From: Merritt Clifton [mailto:[email protected]]
Sent: Thursday, May 01, 2008 3:12 AM
To: Animal Outreach
Cc: Kim Bartlett
Subject: Re: Fwd: 2007 U.S. Animal Shelter Killing Report Card

>>>>>You have Terre Haute (Indiana, I assume) as having a population of 169,000. Its more like 60,000. The Metropolitan area of Terre Haute has about 150,000.

The U.S. Bureau of the Census & the state of Indiana are agreed that the Terre Haute metropolitan area in 2005 contained 168,000 to 169,000 people, with the rounding on the upward side.

>>>>> But the metro area most assuredly killed more than 78 animals. I can almost guarantee you the town of Terre haute killed more than 78. Either way, your data is way off.

First of all, you really should verify what we published before jumping to the conclusion that someone else's reconstruction of the data on another web site is in fact what we published.

The figure we published was 780.

Second, if you had multiplied the rate of shelter killing per 1,000 human residents, 4.6, by the numbers of thousands of human population, 169, you'd have gotten 777.4, which would have told you very quickly what our input figure actually was.

-----Original Message-----
From: Merritt Clifton [mailto:[email protected]]
Sent: Thursday, May 01, 2008 11:49 AM
To: [email protected]
Subject: RE: Fwd: 2007 U.S. Animal Shelter Killing Report Card

>>>>>Ed Boks was simply regurgitating your own data.

Our table was divided by region. Boks completely reconstructed it, in an altogether different manner.

>>>>>page 19, says Terre Haute (not "Terre Haute Metropolitan Area" I suggest that in the future you label your data field accurately) killed 78. Your magazine published 78

If you look closely, what you will see is that the "0" simply didn't print, and that the placement of the "7" and "8" are in the correct columns for 780.

Looks to me as if you just want something to gripe about. Trying howling at the moon.

--
Merritt Clifton
Editor, ANIMAL PEOPLE
P.O. Box 960
Clinton, WA 98236

[Kim Bartlett]

Merritt Clifton left Animal People departed Animal People at the beginning of 2014 to produce his Animals-24/7 blog and website. The board of directors of Animal People, Inc. does not stand by Merritt Clifton's statements and statistics on pit bulls and has never supported breed bans as policy or in editorial statements. Animal People NEVER had a position in favor of breed bans, though Merritt Clifton may have given that impression in what he said or what he wrote in personal emails. At the most, Animal People would favor breedING bans for the reasons that pit bulls and related dog breeds are the most abused dogs, the most likely to be used by dogfighting rings, and the most likely to end up homeless. With a breedING ban, the worst that could happen to a dog would be spaying or castration; none would lose their homes. As an animal rights organization, Animal People does not favor the breeding of any animals for human purposes and does not support the pet industry.