Why Kill Rates are Critical When Charting a Shelter’s Progress

(Note:  “Sample Animal Shelter” as referenced in this post, is a fictitious name for a facility.  The name and numbers are merely being used as an example and are not representative of any actual facility of which I am aware.)

A brief overview, The Maths edition:

If Sample Animal Shelter took in 100 pets in 2011 and killed 90 of them, that would be a 90% kill rate.  If Sample Animal Shelter took in 60 pets in 2012 and killed all 60 of them, that would be a 100% kill rate.  It is misleading to look at the actual numbers of pets killed and compare them by year while excluding the intake numbers.  That’s why I so often refer to kill rates since those figures take into account both intake and outcome figures.

Sadly, there are some pet killing facilities that tout their actual kill numbers, exclusive of total intake, to the media in an attempt to mislead the public.  If Sample Animal Shelter’s director wanted to conceal the dismal comparison between the 2011 and 2012 kill rates – going from 90% to 100%, he would tell the media that the number of actual animals killed in 2012 decreased by 33% compared to 2011 (90 pets killed in 2011 and 60 killed in 2012).  The announcement of this significant decrease fools many people, who don’t know how shelters spin numbers, into believing the people at Sample Animal Shelter worked harder to save animals in 2012 than 2011.  In reality, the facility merely took in fewer animals in 2012 and staff killed every single one of them.

None of this is to say that actual numbers have no value.  Indeed they do.  But when a shelter director publicly claims a decrease in killing over a period of time, it is important to compare kill rates, not actual numbers, since the former tells the complete story and does not require a calculator to draw a relevant comparison.  Most shelter directors either know or should know this.  So too, they either know or should know that the general public, including most of the media, is unaware of the ways shelter kill numbers can be spun to trick donors and taxpayers.

As I often say, if you can’t own it, maybe you shouldn’t be doing it.

19 thoughts on “Why Kill Rates are Critical When Charting a Shelter’s Progress

  1. You always have to compare intake with outcome. Intake is easy, you add up everything that comes to the shelter no matter in what condition. Even if you have dead on arrival, you still have to do the intake. Outcome is a bit more tricky because you have to break it down. Outcome, as example, is adoption, transfer, euthanized in shelter, euthanized at vets office, died in shelter, died at the vets office, dead on arrival.
    So, instead of looking at the kill rate, looking at the life release rate would give you a more accurate picture of the shelter performance.

    1. Live release rate gives no more accurate picture of shelter performance than kill rate. They are simply two sides of the same coin.

      1. How to calculate the life release rate: Adoptions (A) + Transfers (T) + Returns To Owner (RTO) divided by A + T + RTO + Euthanasias (E). Different shelter use different methods for calculations.

    2. I see below where you give the formula for calculating live release rate, but you don’t explain how it gives a more accurate picture of shelter performance than kill rate. You sort of appear to be saying that one should look at the outcome numbers broken down into various subcategories, but the final calculated rate does not provide that information.

      1. What do you consider kill rate? What do you all include in the kill rate? Do you think every shelter will include the same things in the kill rate? All I’m saying is that the life release rate will give you a much better picture about the performance of a shelter.

      2. “All I’m saying is that the life release rate will give you a much better picture about the performance of a shelter.”

        Yes, I know that’s all you’re saying. You’re not explaining how or why it’s better.

    3. “Because animals can die for many different reasons in a shelter.”

      That does not explain how live release rate is a better measure of performance than kill rate. AFAIK kill rate is the kill rate – the number of kills divided by total intake. It takes everything into account, because everything is either kill or not kill. Your formula does not take into account deaths from natural causes or animals remaining in the shelter. Again, how is that better? How is (total live release)/(total live release + kill) a more accurate measure of performance than (kill)/(total intake)?

      1. Basically, in a animal shelter you have two outcomes, dead or alive. The outcome “alive” is pretty simple and straight forward, the animal leaves the shelter alive, end of the story. The outcome “dead” (or kill rate in your words) is much more complicated since you have to include everything even the death that is out of your control. As I stated in another post, our ACO sometimes will pickup a injured animal, takes it to the vet and then the animal dies at the vet office. Animals die of natural causes which you will have to include in the kill rate too (but you actually didn’t kill them).
        Let’s assume the following conversation is taking place:
        A: Hey, my shelter has a 10% kill rate
        B: What? You only killed 10%?
        A: No, we didn’t kill them all, some died for other reasons
        B: Then why do you call it kill rate if you didn’t kill them?

        Sounds confusing to me…..

      2. Who said deaths from natural causes should be included in the kill rate? I absolutely did not say that – I clearly wrote “kills,” not “deaths.” So who is saying that? And again, your formula ignores them completely. They do not get accounted for in any way, shape, or form. They are neither included in the numerator nor the denominator of your formula.

        You still have not explained how live release rate is a better measure of performance than actual kill rate. Ok, so it’s better than some erroneous kill rate? Even if that’s actually how some shelters are calculating kill rates, the solution is not to simply say that the live release rate is “better”; the solution is to calculate kill rate accurately.

        The main point of this short blog post was that you need to know kill rates in addition to the absolute kill numbers because some shelter officials will use decreases in the absolute numbers in a misleading fashion. Shirley did not say that kill rate is the *only* number or the *best* number, and none of what you have written actually addresses her actual point. You simply came in here to derail and mansplain. The fact is that no one number is “better” for an accurate picture of shelter performance – you need several numbers in order to get an accurate picture. I won’t be responding anymore because you are spewing utter nonsense.

      3. You don’t have to respond, that is fine with me. First, my main issue is with the word “kill rate”. I’m not sure if everybody is using the same definition here and how can you tell if a shelter euthanized the animal or killed the animal? The kill shelter sure won’t tell you.
        Second, for No Kill shelter in the No Kill Community (and also for many others) it has become standard to use the live release rate. In fact, many shelter are using the Asilomar Accords for statistics. Remember the formula I provided earlier? Here it is:
        And here is the actual paper you use for your shelter statistic:

        Have a good night :-)

  2. Unless you are going to break down both intake and kill rates by additional criteria, all numbers are somewhat bogus. There are some states where a shelter is required to add injured/dead on arrival wildlife into their “kill rates”. That is not fair and does not present a true picture of the shelters’ success or presumed failure. The other biggest mis-interpreted number is intakes. Here is an example: my group received a dog from out of state, the dog was subsequently transferred to one of our foster homes in another state. The dog was then moved to a different foster home in yet another state. So here are the numbers for ONE dog:

    MO – intake
    IL – intake
    MI – intake
    IA – intake and eventual adoption

    This dog was marked as an incoming dog FOUR times! Since so many shelters transfer dogs in and out but have no official requirement (in some states) to add transfer numbers into their reporting criteria, you can easily see how the animal rights and pet overpopulation folks get some very inflated numbers. Untill the reporting process is either uniform or streamlined we should not rely on the numbers given for a particular group.

    1. Taxpayer funded facilities have an obligation to inform taxpayers how many animals they take in, kill, transfer, adopt, etc. You are clouding the issue by referring to pets transferred between private rescues.

      1. The first group was not private, it was an animal control facility. In addition, the transfer of this dog may or may not have been counted as an adoption, just an intake.

        Taxpayer funded facilities only report what they are required by law to report and that was my point. They are normally not required to report intrastate or interstate transfers on their reporting forms, nor are most required to record domestic animal euthanizations vs wildlife or tally the numbers of animals brought in by owners for low cost euthanizations. Another number never tallied by most is the number of feral cats euthanized vs owned animals. Without specific data you get skewed conclusions.

      2. The pet would have been listed as live release, even if the facility didn’t break down adoption vs. transfer.

        Many taxpayer funded facilities report, even if not required by law. Maddie’s Fund grant recipients are required to report. Particularly loud and annoying taxpayers can raise hell over their local facility’s failure to report and force them to do so. Or so I hear…

        On Mon, May 13, 2013 at 11:20 AM, YesBiscuit!

    2. This is easy. It is a live intake that is a “transfer in” from another group. Then it is a “transfer out” to another group. One in, same one out. Even Steven. That’s all there is to it. The one tricky part is when a group transfers out to a kill shelter. The transfer out should be to an organization with the near certain probability that the animal will be placed (or humanely held).

  3. Thank you for this post again Shirley. Shelters spin numbers every way they can. As you know we are trying to work with Metro Nashville Animal Control. They are spinners. In an statement released by the Director in January – the spin is so evident and even the media interviewer didn’t take her to task on her statistical BS. Here’s the article:

    As Shirley says – here’s the money quote: “The numbers are better,” she said. “Our intake was lower, which automatically means our euthanasia rate was lower than in 2011.” WRONG.
    Taking in LESS animals does not automatically mean that their Kill Rate is lower. Infact – Kill Rate has NOT gone down at Metro Animal Control.

  4. If I’ve learned nothing else, it’s that if someone is looking to spin numbers, there are plenty of ways to do it.

    I don’t necessarily like using straight kill rate either, because there are a number of ways that can be done, both good and bad.

    If a shelter is doing a very good job of providing outreach for instance, or ridding themselves of very anti-pet laws such as mandatory spay/neuter, low pet limits, etc and lowering shelter intake by keeping more animals in homes, then they are, indeed, saving more lives by keeping them out of the shelter in the first place — even they aren’t actually “saving” more animals once they get there. No kill is math, # out safely has to be at or near intake or you lose. So lowering intake is one way to get there.

    On the other hand, I don’t like that some shelters have simply achieved better kill rates just by limiting intake. A shelter that took in 5,000 animals a year before, and saved 2,000 of them, if they rid themselves of a city contract and save 2,000 the next year, but the other 3,000 get killed somewhere else, there are no net lives saved. So that’s no good either.

    I guess what I’m saying is that there are ways to lie regardless of the numbers you use….but the end goal should be saving more lives, and saving all of them (whether by saving them, or keeping them out of the shelter in the first place), without trying to fudge numbers to get to a target number.

    1. I agree with you. The devil is in the detail, so to speak. As example, ACO brings in a dead animal, you still have intake and outcome. Another person is bringing in a litter of kittens, 2 of the kittens die after 1 hour and again you have intake and outcome. ACO is picking up injured dog from the street and is taking the dog to the vet, dog dies 2 hours later at the vet office, the shelter still has intake and outcome.
      It is not always that simple or just black and white.

Leave a Reply