Best Friends
No More Homeless Pets Forum
January 5, 2004

What Can Statistics Do For You?

Richard Avanzino
Richard Avanzino

Statistics and definitions: What can they do for you? Richard Avanzino of Maddie's Fund will answer your questions about the hows and whys of gathering data and establishing definitions, and using it to formulate plans for your organization and community.

Introduction from Richard Avanzino:

Statistics may be a ho-hum topic to some, but, in my opinion, tracking numbers is key to the success of any animal welfare organization. Statistics are navigation devices -- they tell you where you are, where you've been and what you need to do to get to where you want to go.

The most important statistics for any animal shelter to track are: the number of animals coming in, the number of animals adopted, the number of animals transferred or returned to owner and the number of animals euthanized. These numbers tell you how well you're fulfilling your lifesaving mission.

Within the euthanized category, it's important to define the animal's condition. Was he healthy/adoptable? Treatable? Non-rehabilitatable? This is essential for several reasons.

Definitions help guide organizational resource allocation. Animals in different categories have different needs and require different resources (e.g., marketing for healthy animals as opposed to medical care for sick and injured treatable animals).

Definitions enable an organization to establish measurable, quantifiable goals. For example, it's hard to set the goal of saving all healthy/adoptable animals without defining healthy/adoptable.

Statistics and definitions give an organization an objective standard for measuring results. They provide the tools to determine how well your organization is doing compared to others or compared to your own accomplishments last year. As statistics and definitions are published, they provide the organizational transparency and measuring stick the public is looking for.

In sum: statistics and definitions help create organizational efficiencies, define organizational goals and priorities, save lives, enhance image and build public support. It doesn't get any better than that.


Creating common definitions and making it easy for shelters to share
Collecting data on spay/neuter from the community
Defining what is adoptable
Formulas to determine unowned animals in a community
Verifying the accuracy of stats that animal welfare organizations provide
What are hurdles to getting groups to share stats and how to overcome?
Getting shelters to honestly report and not mislead with stats
How many litters can one female really produce?
Keeping track of adoption stats
Getting accurate information if you believe a shelter is not sharing correct stats
Starting to gather stats when your community keeps none
What stats should a shelter track?
Avoiding common mistakes in gathering data
Should we use "per 1,000" stats instead of categorizing animals?
Tracking stats about why animals end up in shelters and developing programs to address

Creating common definitions and making it easy for shelters to share

Question from a member:

I volunteer with a developing coalition of animal shelter and rescue groups - communicating with over 75 animal welfare organizations in nine counties. We are in the early stages of community needs assessment - gathering statistics from the primary agencies in our area. Most organizations have been cooperative and forthcoming with basic statistics, intake, adoptions, transfers, and euthanasia. But none of the groups is tracking types of intake and reasons for euthanasia thus making assessment a difficult process.

Meetings with shelters and rescues and references to statistic gathering publications have generated some curiosity. But most agencies site lack of staffing resources, show an unwillingness to change procedure, and do not have the software to support recording specific information.

How can we, as a coalition

- encourage cooperation amongst organizations
- help them to create common definitions/categories for animals (adoptable, rehabilitatable, etc.) similar to what they've done in the Denver area
- what is the most time efficient, most cost effective, accurate way for shelters to record more detailed statistical information?

Response from Richard:

To convince organizations to collect, provide and monitor statistics, they have to understand why they need to make the effort to get the data.

To me, collecting statistics is the only way to look back on what an organization and a community have done to fulfill their purpose in helping animals in need.

Statistics are the only way to understand where you are now in achieving your goals. This information would seem to be an essential pre-condition for developing a plan that measures how you're going to get to where you want to go.

It's also founded on the concept that the status quo isn't good enough, that lives are at stake and we have to do a better job.

If these principles are agreed to, then it's obvious that the data needs to be gathered, provided and used.

You asked about helping the groups create definitions. I would suggest using definitions already in common use. Many groups use the Maddie's Fund definitions and categories. A good example is San Diego County's Department of Animal Services. Denver's categories are very similar to Maddie's.

By using already established and widely used definitions, you have a means of comparing your data to that of others on an "apples to apples" basis. If each community comes up with its own set of definitions and categories, it will be as if each community is speaking a different language.

In terms of shelters finding a time efficient, cost effective way to gather statistics, I can suggest a few different approaches (some with more accuracy than others).

We're fortunate that nowadays there are several computer software programs that can make detailed tracking pretty easy. Even if the wording in these programs doesn't dovetail exactly to the category you're tracking, you can easily make the translation. For example, "vicious dog" translates into "non-rehabilitatable" and so on.

Another way to short-cut the process is to carefully apply the definitions to the shelter population for three months, then extrapolate the information to a twelve month period.

And finally, you can ask employees who have been on the job for many years to make their best guess in placing dogs and cats in the various categories, using the definitions you provide them.

Collecting data on spay/neuter from the community

Question from Julie:

Much of the data my organization hopes (intends) to track is not gathered "in house" For example, we want to track the typical costs of spay/neuter for low-income pet owners in several communities, and the impact on shelter intake that a particular spay/neuter subsidy program might have. While we can carefully define the data we want, actually getting that data is whole 'nether adventure. In your experience, what are the keys to encouraging data sharing among separate municipal animal control, humane organizations and veterinary businesses?

Response from Richard:

The first thing I might do to give your project more authority is write an official "White Paper" that describes what you're trying to accomplish, why you're trying to do it, who is supporting it, etc. Each business model you target - veterinarians, municipal animal control or humane society, has different interests and agendas. Your paper should aim to convince each group that your project is credible, is beneficial to the people and pets in the community, and is helpful to them. Ultimately, you'll want to look at each business model and figure out what technique, information or leverage you can use to convince the various groups to get on board.


Veterinarians generally fall into two categories: 1) veterinarians who are very interested in the animal welfare agenda and active in community spay/neuter programs and 2) veterinarians who are interested in animal welfare, but not as directly engaged.

In general, veterinarians are reluctant to give out statistics about their practice for all of the reasons (time, privacy, etc.) mentioned in first question of this forum. These concerns are heightened by the fact that, as a for-profit business, doctors have to keep their eye on the bottom line. While they may want to help the cause, it will take time and money to provide statistical information, and unless they are philosophically aligned with the animal welfare agenda, they may be very reluctant to do so. Veterinarians less active in animal welfare may only be persuaded to share their statistics by providing them with monetary compensation in the form of spay/neuter surgeries or other business or reimbursement for time spent collecting and providing the data requested.

Many veterinarians are concerned that statistics will be used against them. This worry can be alleviated by assuring the doctors that information collected will only be published in aggregate and not by individual practice.

As a general rule, veterinarians have a greater trust and comfort level when approached by their peers. We have been very successful in working with veterinarians through veterinary medical associations.


As government agencies funded by taxpayer dollars, municipal animal control organizations should willingly provide their statistics. If the information is not readily available or the shelter is reticent, you might have to get access through an elected official who oversees the agency's budget.


Most humane organizations are eager to publicize their spay/neuter efforts—the number of surgeries they perform, charitable dollars spent, etc. This kind of information builds support and buy-in from the community, support badly needed to fund their services.

Humane organizations are also generally willing to share their intake numbers. Intake is far less sensitive than euthanasia.

If several organizations agree to provide data but a few holdouts remain, many times peer pressure will get the reluctant parties to come on board. When one director says to the other, "everyone else is here but you. We're trying to do a comprehensive report and we need your information to make it complete," that can carry a lot of weight.

Defining what is adoptable

Question from a member:

Groups often disagree on the definition of adoptable. For instance, some deem feral cats or specific breeds unadoptable. How can we get everyone on the same page about the definition of what is and is not adoptable since this can have a dramatic impact on the numbers? Is this a pretty contentious issue in most communities?

Response from Richard:

We have heard that some organizations have found that trying to define adoptable can be contentious.

To us, it doesn't seem that categorizing animals based on health and rehabilitation status should be so difficult.

Maddie's Fund uses three categories: healthy/adoptable, treatable and non-rehabilitatable.

Healthy/adoptable is defined as: Animals eight weeks of age or older that, at or subsequent to the time the animal is impounded or otherwise taken into possession, have manifested no sign of a behavioral or temperamental defect that could pose a health or safety risk or otherwise make the animal unsuitable for placement as a pet, and have manifested no sign of disease, injury, or congenital or hereditary condition that adversely affects the health of the animal or that is likely to adversely affect the animals health in the future. Healthy/adoptable animals may be old, deaf, blind, disfigured or disabled.

This definition seems pretty straightforward. When it comes to the treatable and non-rehabilitatable categories, I believe there are a few gray areas where people might have an honest difference of opinion. But I also believe these gray areas are very small.

I think everyone agrees that a vicious dog is a public safety risk and should not be placed. I think everyone agrees that a 20-year old cat suffering from terminal cancer and in pain is not a good candidate for rehabilitation.

One organization, San Diego County's Department of Animal Services, has taken the adoptable, treatable and non-rehabilitatable categories and broken them down into sub categories to help them with the all important task of applying the definitions and explaining what the terms mean to them.

I think we can also agree that when definitions are trumped by ordinances the law should be followed. For example, breed specific legislation sometimes prohibits animals like pit bulls or feral cats from being put up for adoption. While we may not agree with this kind of legislation, it should be honored to reach community consensus until it can be changed.

It is extremely beneficial for the animals, the people who work for them, and the community as a whole to find common ground. We need to reach consensus when possible and work out our differences amongst ourselves.

At the end of the day, I believe we can agree on the vast majority of issues with no difficulty whatsoever and the few cases where we don't agree should not divide us or take us from our united commitment to save the lives we are entrusted with.

Formulas to determine unowned animals in a community

Question from Carla:

I have read arguments to the effect that unowned animals contribute far more to the population problem than do owned animals. In efforts to promote more shelter space, adoption facilities, and above all, affordable spay-neuter, it would help to know how many animals in total exist per capita. I can find formulas for owned animals per capita but have no clue about unowned animals (cats and dogs, that is). Is there a way to make a plausible estimate for a given state, county or city?

Response from Richard:

For starters, I would say the population of unowned dogs in most communities in the United States is relatively small, although there are a few regional pockets throughout the country where larger populations do exist. Animal control departments have historically kept free roaming and unowned dogs in check due to concerns about rabies.

Cats are a very different matter. There are populations of unowned and free-roaming domestic cats as well as true ferals. To get a better handle on these numbers, I called Animal People editor, Merritt Clifton.

"Based on road kill data plus shelter intakes, I don't think the current feral cat population is more than about 8.5 million. Five million would be my low-end estimate. Split the difference and it might be 6.75 million, but the local distribution will vary from almost none to 40-50% of the total number of cats. So anyone working on this issue really needs to do local-level research before publicizing an estimate."

In your opening sentence, you imply that the unowned population potentially creates a large overpopulation and/or sheltering problem. But does it really? Maybe that depends on the definition of overpopulation. Does overpopulation mean too many born? Too many in a community? Too many in a shelter? It's an interesting discussion for another time.

Verifying the accuracy of stats that animal welfare organizations provide

Question from a member:

We are a new foundation and we have looked at existing statistics from one organization on our own island and another neighboring island. The statistics are like night and day. On one island, Animal Welfare claim to have "picked up" hundreds of dogs and cats in 2003. There is no mention of how many were euthanized. On another island, where the population is not much more than the other island, the stats show that an alarming number of thousands of dogs and cats were euthanized in 2002! How can we verify the accuracy of statistics given by animal welfare organizations?

Response from Richard:

It sounds like the information you're working from is pretty vague and incomplete. I would suggest a letter to the Executive Director that asks for a response to a detailed list of questions. You want to know animal intake numbers per year and by species. You want to know animal outcomes-deaths, adoptions, transfers and returns to owner. You definitely want to get information on shelter policies as they relate to strays, ferals, and owner surrenders. For example, it is conceivable that one shelter has a restricted intake policy - maybe on feral cats or owner surrendered animals. No-kill shelters limit their intake to the numbers of animals they can care for and re-home. Obliviously, policies like these will radically affect the statistics.

Once you have a more complete picture, you can contrast your information to other communities through publications like Animal People, which has a rough yardstick for comparative purposes. Maddie's Fund also collects numbers and statistics and we've formulated our own internal guidelines for making community-by-community comparisons.

If the organization is unwilling to provide the information you're seeking, if the information provided doesn't make sense, or if you feel the agency is manipulating the numbers, you'll no doubt feel uncomfortable investing in this organization.

In our own grant giving, Maddie's Fund starts with the assumption that animal welfare organizations are honest, reputable and devoted to the well being of their animals. Until proven otherwise, we operate with this assumption as an extension of our own core values of honesty, integrity and mutual respect.

The structure of our grant giving helps minimize problems. We require collaborations of animal control agencies, no-kill organizations, traditional shelters and rescue groups. This collaborative structure provides some checks and balances as the different groups look over each other's numbers. If there are multiple groups on the island, and you can get them all to submit their statistics, the peer review may help cut down on spin heavy reporting.

Comment from Christine, DVM:

There is NO accurate data concerning the numbers of feral cats, stray unaltered cats, reproducing cats in the US. That is a fact that is certain. ALL numbers are hypothesized, so any "calculation" of a hypothesized number is no better than a guess. Everyone becomes an expert because anyone can guess.

It is NOT certain that a female cat can produce three litters in one year. That is merely a potential but not reality. Quite soon there will be a study published in JAVMA that was performed by a wildlife biologist who has no vested interest in proving anything about TNR, euthanasia, suffering, etc. regarding free roaming cats. The study was begun by a PhD student whose senior advisor continued the study in her absence. The numbers are interesting and should be seriously considered.

Basically, the study looked at 10 managed feral colonies. Three control colonies were allowed to breed for three years before TNR was introduced. There is much important preliminary data coming from this study but of greatest interest and probably of greatest use in fundraising is the fact that most females only have ONE live-birth litter per year, despite optimal feeding, vaccinating, shelter and climate (North Carolina). The study was designed to make all conditions the best for maximizing reproduction in the colonies. The mean litter size per year was 1.3. It was very unusual for females to birth two live litters per year. Yes, they might have been pregnant more than once per year, but live births were NOT the three per year that some folks say is a "fact". Now consider that these numbers are from OPTIMAL conditions in a moderate climate and apply them to most colonies elsewhere with less than optimal conditions. It is highly unlikely that cats with anything less than optimal living conditions will produce more than the 1.3 litters per year average.

More data showed that unaltered females generally DON'T survive seven years but more like 3-4 yrs. But what is MOST important to reconcile that preposterous number (420,000 births prevented by altering one female) is that almost 50% of kittens die before 3 months of age, and merely 70% die by 6 months. Remember this is still with all conditions optimized. That means about 3/4 of kittens NEVER BREED because they die before they reach breeding age. Perhaps this explains why we're not over run with cats all over this country? These numbers make perfect sense when compared to other wild carnivores. It is nature's intent to produce more offspring to increase the chances that one will survive to perpetuate the species. Sea turtles have hundreds of eggs with hope that one will survive to breed. Cats are no different.

So if you want to play with numbers, let's go with one female cat bearing one live litter per year, of the average four kittens per litter. Because of kitten mortality, only one will survive beyond 6 months and potentially breed. Chances of being male/female are probably 50/50 (who knows?), so we do NOT get 420,000 cats in 7 years from one female. Seems far more realistic and logical than one female bearing 12 kittens per year half of whom are female bearing three litters per year and so on. That level of population growth and survival is simply not what people observe in colonies anywhere.

When it comes to fundraising and donor appeal, kitten mortality numbers are the numbers that we use. The fact that over half the kittens born will perish in the "wild" is a fact and is a powerful point to make.

More interesting data from this study (presented at the American Veterinary Medical Association Animal Welfare Forum in Chicago in November, 2003): vasectomized males have the same shorter life span of intact (tom) males. Castrated males have the same longer life span of spayed females. Spayed females live longer than intact females. Altered cats gained weight and were overall healthier than the unaltered cats.

The percent of a colony needed to sterilize in order to stop colony size increasing is MADE UP. It's simply a guess.

Number of feti in pregnant cats is higher than the number of live births.
Most groups are not seeing colonies of 50/50 females. More like 60-70% females, so that skews the projection of stats produced from "if half the kittens are female..."

Good for you for helping your community and good for you for wanting accurate information. We believe that the best measure of results is not some preposterous projection of births prevented, but of the tangible, immediately evident numbers of spays/neuters that you're performing. Every animal altered will prevent deaths. Richard Avanzino is exactly right - Start bragging about what you do, the facts that you can prove. You are working hard, battling looming obstacles, but remember there are many people - and people with money - that are out there looking for a group like you to support. Many people are coming to realize that shelters and sanctuaries are not shutting off the flow of animals being produced in this world.

That reminds me of another interesting set of numbers: One presenter at the meeting showed a sanctuary in Brazil. It housed 1200 cats and was filled soon after opening. Attrition rate was low because the cats received great care and medical attention. Cost for maintaining the sanctuary is $25,000 PER MONTH! I couldn't help but imagine the number of spay/neuters that could be performed for that amount of money. I realize that some individual passion is in sanctuary/shelter/adoption, and I do applaud the efforts, but I fear that we are bailing a sinking ship and more folks need to help plug the leak rather than bail the water with a teaspoon. That is the opinion of myself, not of my organization.

My deepest apologies for not properly attributing the science to the source - I am buried under a stack of overdue work. I can come up with the names to those who cannot await their publication, but I believe the studies will be published in the Feb issues of the JAVMA.

What are hurdles to getting groups to share stats and how to overcome?

Question from Michelle:

We recently formed a coalition. While hosting a Super Adoption to generate interest and excitement in our area about this effort is a major goal for the first half of 2004, another one of our first-year priorities for the second half of 2004 includes assessing our community by gathering relevant statistics about several counties within our geographic area. To prepare us for the road ahead: What are some of the biggest hurdles other groups have faced in collecting statistics and how can we circumvent these hurdles as we begin this process? Also, aside from including shelters in this process, does it also make sense to have representation from the county government at a higher level to encourage buy-in about the results?

Response from Richard:

For starters, I would offer a word of caution. You mention gathering statistics for several counties within your geographic area. That may be a bit ambitious.

In order to have a sufficient data base, you'll need to get the cooperation of the major players within each community - animal control, traditional shelters and rescue. The more organizations, the greater the challenge: consensus is harder to reach, organizational differences are harder to overcome and personality conflicts play a more significant negative role.

I would start with just one county, the one with the greatest interest in collaboration among the major groups. Once you've collected the data from this core coalition, it will be easier to expand one county at a time.

As you get ready to ask the different agencies for their statistics, remember the first thing everyone will want to know is "What's in it for me?" Be prepared with an answer, because as you suggested, there will be a variety of hurdles:

Some shelters, especially smaller ones, may not keep statistics. There may be resistance to doing something that's never been done before. Other shelters that keep statistics may not keep them the way you want them and may see no reason to change their methods just because you ask them to. Many animal control and traditional shelters fear the statistics will be used against them (their death numbers will be published and make them look bad). Some shelter staff will feel their statistics are none of your business.

Shelter workers that believe they are already under staffed and overwhelmed with lifesaving functions may find gathering statistics too time consuming.

In Maddie's Fund projects, we ask that shelters categorize animals as healthy/adoptable, treatable and non-rehabilitatable. Shelter managers may not see the need for these categories, agree with the categories or find it useful to apply the categories. To get buy-in and cooperation, the agencies will have to believe it's in their self interest and the interest of their community to participate in a collaborative approach to solving pet overpopulation. Each group has to believe that by working together, they'll have a bigger impact and greater success than working by themselves. When groups feel that what you're asking for is ultimately going to help them and save more lives, the hurdles will be easier to overcome.

In response to the second part of your question, I would wait to approach the County until you have a cohesive collaboration of animal organizations. You need to show that there is a united front working together to solve a community problem before you ask government to get involved.

Getting shelters to honestly report and not mislead with stats

Question from a member:

Our coalition has been gathering statistics for about a year with our statistics overwhelmingly showing that cat overpopulation is a huge problem. Some areas face an over 50% upwards of 85% euthanasia of cats, with a large number falling into the non-rehabilitatable definition and very few fall into the adoptable and treatable definitions. We do know that shelters euthanize cats for upper respiratory, while that is truly not a non-rehabilitatable condition. It is treatable.

How can we get our shelters to be more honest with themselves and to understand the value of keeping these statistics by defining animals correctly? At the same time, the whole idea of definitions and terms seems very subjective. It seems by keeping these statistics, a grant could be written requesting monies on the basis that "according to our statistics, if we would have had funding to treat this many animals, we could have saved them". How can we help them to start seeing the value in reallocating funds, first for spay/neuter since that will reduce their intake and allow them to have the time to actually write grants and implement programs to treat treatables, and also offering a TNR program for ferals?

Response from Richard:

There seems to be two very different questions here. One: how do we get the shelter to honestly report? Two: how do we get the shelter to start spending more money on spay/neuter? I can't emphasize enough how critically important I think it is for organizations to be honest about their activities and accomplishments.

Statistics are a big help in that regard. Organizations that keep accurate statistics have a truthful understanding of their progress, past and present. They also have a handle on what they need to do to get to where they want to go in the future. By manipulating numbers and definitions, an agency interferes with its ability to make progress in saving lives. When statistics are manipulated or definitions misused for public relations purposes, the organization is just shooting itself in the foot.

It's not uncommon for organizations involved in social causes to try and avoid scrutiny by misrepresenting information. These organizations usually don't get away with it for very long, and shelters are no different. Once the community finds out the truth, a trust is broken that is hard to repair. This will affect the shelter's ability to raise funds and save lives in the future.

The real answer in my mind is to constantly promote the use of common definitions and encourage universal acceptance. Once the definitions become widely accepted and comparisons are made, it will become clear which shelters are conning the public.

As for allocating more resources for spay and neuter: Maddie's Fund invests in a two-pronged approach that supports both spay/neuter and adoption efforts. Yes, money should be spent on spay/neuter to reduce the problem for tomorrow, but we can't ignore our responsibility to the cats and dogs that have already been born. Their lives need to be saved today.

How many litters can one female really produce?

Question from Christine:

Several friends and myself have started a spay/neuter assistance group. Our main goal is spay/neuter and we are trying to fill the gap between the truly poor, which are served by Maddie's Fund and the working poor who are unable to spay neuter their cats. When our resources become more plentiful we plan on trying to assist large dog owners to help offset their costs. We have been emphasizing working with feral feeders who have limited resources, however that is only because the word has been slow to get out that we will subsidize/pay for altering owned animals.

We have been doing very well at fund-raising and are ready for our first fund-raising letter and publication of our first brochure. We want to be accurate about our statistics and I've heard many numbers bounced around about how many animals are prevented from birth by a single spay/neuter. I don't want to insult anyone's intelligence bating about huge numbers if they are not statically accurate, but I don't want to undersell our accomplishments (120 spay/neuters in just over two months). Do you know where I would find reliable figures on how many lives are saved by spay/neuter?

Response from Richard:

Christine, I applaud you on two counts. First, you're doing a great job in getting a lot of underserved animals spayed and neutered. Second, you're seriously concerned about reporting accurate and reliable information in your fundraising appeal.

In terms of stating how many animals are prevented from being born as the result of a single spay/neuter surgery, the following formula has been commonly used for decades:

One female cat and all her offspring can produce 420,000 cats in seven years...

Personally, I don't think there's one universal formula that can be applied nationwide and I doubt that the numbers above represent reality. So many variables come into play when you try to make this kind of computation: Are the cats indoors or outdoors? What is the health of the cats breeding? If they are free roaming, what are the existing environmental factors such as weather, disease and predators? What is the responsibility level of the various caregivers in avoiding pregnancies?

The one thing you can say with certainty is that one cat can produce three litters per year. If you disclose how you arrive at a calculation beyond that, that is, if your literature explains the assumptions - "assuming each cat had an average litter of four, and half of the cats were female, and they lived to seven years of age" - then you are using an approach that many people find acceptable.

If it were I, however, I think I would try to sell the program in a different way. Why focus on the hypothetical when you can talk about the great work you've already accomplished? "We're a start-up group that began only _____ ago. In our first two months, we've already spayed and neutered 120 animals. We've gotten the support of the veterinary community; doctors and veterinary hospitals are supporting our efforts..." List out all of your successes, set out numerical goals for the future, enlist help in reaching those goals and promise to report back on how close you've come to reaching the goals in your next letter and brochure. Good luck!

Keeping track of adoption stats

Question from Billy:

You mention that the most important statistics for any animal shelter to track are: "the number of animals coming in, the number of animals adopted, the number of animals transferred or returned to owner and the number of animals euthanized. These numbers tell you how well you're fulfilling your lifesaving mission."

But what about the number of animals any one individual adopts from the shelters? There are people who continuously adopt dogs for various reasons such as: keeping them for family dogs, and when they don't work out the dog is returned to the shelter or dumped, then the family gets another one, only to repeat the cycle, sometimes getting & dumping many dogs ad infinitum; or other people who are willing to pay the required fees to take a dog for use in illegal dog fighting, and they are continually allowed to do this, as there is no effort or requirement on the shelter's part to follow up on individual adoptions, or to even keep track of how many animals a person adopts out. Any suggestions based on your experience with shelters on how to implement a program to deal with this?

Response from Richard:

I believe that shelters with adoption programs want to do what's right by the animals. But if an adoption program results in a revolving door where animals constantly go out and get returned, the shelter is not fulfilling the intended purpose of saving lives. When shelters have to place the same animals over and over again, precious resources are wasted and the shelter's ability to reduce overall deaths is seriously impaired.

Billy, I think good communication, screening and tracking are the keys to successful adoptions.

A shelter is off to a good start when employees are taught how to provide high-quality customer service, use good active listening skills, and are provided with guidance from experienced staffers that will help them detect problem adopters.

After adoptions are made, shelters need to track the successes and failures of their placements. If there are significant problems with inappropriate adopters, the shelter will need to set up mechanisms to correct and reduce these incidents.

There are a variety of solutions available, depending on the severity of the problem and the type of shelter operation (e.g., animal control, traditional shelter or rescue group). For example, if the worry is over the number of animals placed per individual, this can be monitored by computer software that sorts by name, address and zip code.

Any shelter worth its salt will have thought about the issues you've raised and put some sort of preventive measures in place. It's irresponsible not to address these concerns and shelter operators know it.

Getting accurate information if you believe a shelter is not sharing correct stats

Question from a member:

If you feel a city or county animal shelter's record-keeping statistics are suspect for whatever reason(s), how can you best circumvent the normal channels so you can get accurate information?

Response from Richard:

That's a tough question. I think it's extremely difficult to get accurate information when you don't have buy-in from the agency responsible for creating it.

I suggest you try establishing a communication link with the shelter operator. Express your thoughts in a non-threatening and non-accusatory manner as to why you think the information is suspect. I would also try to offer assistance in addressing the issues the shelter may put forward.

If lack of staff time is the shelter's biggest problem, maybe you can volunteer your own time or that of others to help gather reliable statistics. If the lack of a trustworthy tracking system is the problem, maybe you can help find or set up a computer program. If categorizing animals is the issue maybe you can provide some concrete examples of other shelters that have successfully applied categories and definitions, such as San Diego County's Department of Animal Services.

If none of this helps, and you believe serious misrepresentations are taking place, you can try going to the elected officials to see if you can use their influence to get accurate data. You might also interest the media in doing a story, but you'll need to present reporters with some concrete information. Nobody wants to spend time on a wild goose chase.

Starting to gather stats when your community keeps none

Question from Heidi:

I am director of a sanctuary that has been operational now for about 18 months. We have been doing very well and we have always talked about assisting our local town, which has a population of only 1900 people, with a no kill policy for their stray and unwanted animals. Under the current system any stray or unwanted animals are shot, on the spot, by police. We would like work with the local officials in order to stop this practice and we feel we are in a position now to get statistics together to see what size animal population we would be assisting.

I realize that knowing the numbers of homeless pets that are in our community is vital information that must be studied by our board of directors in order to successfully assist with this task. In a town that has no records and has no animal control where do we start to gather the numbers. Any advice you would have for us would be appreciated.

Response from Richard:

The situation you describe sounds troubling. It's laudable that your animal welfare organization wants to help in such a dire circumstance. It concerns me greatly, however, that the community and its elected officials haven't accepted the responsibility of providing local residents with humane animal control services.

Here's my worry about your getting involved: most organizations can't do their own work and the government's job and perform both at a high standard.

It's a very tough task for a charity to carry out its mission and take on the added responsibility of government using philanthropic dollars. It's a really bad situation if the job isn't done right and you lose economic support from your own contributors as a result. Bottom line - if you can't do the job well, don't do it.

You mentioned you wanted to help your town implement a no-kill policy. Given everything I've already said, this seems to be an especially ambitious goal.

Getting back to your question about gathering statistics: as a starting point, I think I would ask the police how many animals they shoot in a given week or month. Then I would ask, if provided with a humane alternative, what percent would they be willing or able to transfer. You want to get the transfer number because no doubt some of the animals killed are seriously injured and diseased. And I assume the sanctuary doesn't want to accept the responsibility of killing animals it can't save.

What stats should a shelter track?

Question from a member:

How detailed should records be at a public shelter? For instance, should they separate statistics for dogs and cats adopted? Is it usual to separate the species in quarterly and yearly reports? What about separating out puppies and kittens from dogs and cats? What about euthanasias requested vs. just euthanasia? Should they keep track of why animals are entering in the first place (stray, cruelty, owner surrender and why)?

Response from Richard:

The first question to be asked is: Why is the information being collected? Statistics should be used to measure progress towards goal attainment and to identify problems so that modifications can be made accordingly.

From Maddie's point of view, the primary goal of any shelter should be saving lives, and to do that a shelter needs basic statistics: number of animals coming in, adopted, transferred, returned to owner and deaths by category (healthy/adoptable, treatable, non-rehabilitatable).

A public shelter also needs to collect data that is relevant to what that agency has been tasked to do for the community. For example: if dog bites are a local problem, the shelter will want to track the number of incidents, location of incidents, breeds involved, dog bite histories, commonalities between incidents, etc.

I personally wouldn't be interested in tracking owner requested euthanasia. To me, the fact that the owner requested the death is irrelevant. I would want to know only if the animal could have been saved and placed in a loving home. Therefore, all I would be concerned about was the animal's status when he was put down - healthy/adoptable, treatable, or non-rehabilitatable.

Bottom line: I think every shelter should track the basics, but if there's a reason to go beyond because it serves an organizational or community purpose, do it!

Avoiding common mistakes in gathering data

Question from Julie:

After we clarify our definitions, what else can we do to avoid common mistakes in gathering and tracking data? And what ARE those mistakes?

Response from Richard:

Once you've clarified your definitions, I think you should look at:

current versus past applications of definitions, and how to reconcile any differences or discrepancies between the two. (One group discovered that animals they had counted as domestic cats in the past were really feral cats. Once they realized the mistake, they had to shift the cats into a different category and reshuffle their numbers.),
whether there's a monitoring system in place to verify whether everyone in the organization is applying the definitions uniformly, how other community shelters are using and applying the definitions.

I don't think you can assume that individual employees or agencies are on the same page when it comes to definitions and categories, even if they're all using the same ones. People have a tendency to interpret situations according to their individual interests and values. When I was in San Francisco, all of the organizations used the same definitions. Even so, animal control applied the healthy/adoptable definition to their own animals with a great deal of leniency to ensure the greatest possible numbers of animals were transferred to our shelter, where they were guaranteed a loving home.

The only solution is to meet frequently with staff and/or other agencies to go over the definitions and hammer out the various issues that people might have. Some of the most contentious issues these days are over behavior. Is a food protective dog who snaps when you approach his dish a treatable or non-rehabilitatable animal? What if he bit somebody in the past? Would it make a difference if it were a minor or severe bite? Does it matter if he's a Chihuahua or a Lab? Hopefully, the airing out of differences will eventually lead to consensus.

Comment from Nonie:

Although I've never used it myself, my understanding is that PetWhere software does have the ability to track PEOPLE as well as ANIMALS. It's even possible to give potential adopters some kind of "flag" so that when the adoption clerk pulls them up, they are notified that that person has some sort of issue. (An active cruelty investigation, for example.) I'd imagine that it would be pretty easy for some computer savvy person to make the software automatically flag a "repeat adopter" after a certain number has been reached.

You can read about (and download a sample) of
PetWhere from the ASPCA website. It is free to any non-profit animal shelter. It is already in use at over 1,200 shelters (according to the ASPCA).

Should we use "per 1,000" stats instead of categorizing animals?

Question from a member:

I've been to a number of conferences around the country and have talked to a lot of people about your (and the Hayden Bill's) categorization of euthanized animals. I've found that people in animal welfare almost universally have difficulty with it. Most of the objections cite examples of shelters distorting their adoptable numbers, either deliberately or not, in order to create a particular impression. And the weird thing is the distortion can take place either way. Minimizing their numbers to convey that they're only killing sick animals, or maximizing them to convey a worse situation than actually exists in order to plea for more help, finances, etc.

Most people support the total animals killed per 1000 population method of counting. It's objective, whereas the other method is so easily distorted by subjective viewpoints, and problems of interpretation. Wouldn't it be better to establish the per 1000 method as the universally accepted system of counting, rather than having competing methods?

Response from Richard:

Yes, I've heard that some shelters distort their adoptable, treatable or non-rehabilatatible numbers. But you're basically arguing, "If they're going to manipulate it, why track it?"

If you're trying to gauge a community's life saving progress, tabulating the deaths per 1,000 has benefit. But if you are a shelter operator who is trying to reduce the killing, you absolutely need to categorize the animals and know the status of the animal at the time of death (healthy/adoptable, treatable or non-rehabilitatible). If you don't know what the animals needs, you can't fashion a workable solution, for example healthy animals need a home, treatables need medical or behavioral intervention, non-rehabilitatable can't be saved.

I'm probably repeating myself, but, in my opinion, the problem of statistical distortion will eventually be corrected in two ways: 1) through broad acceptance of standard definitions and a generally uniform application of the definitions, 2) by increased opportunities for comparative shelter by shelter analysis.

As more and more shelters (e.g. Tompkins County SPCA, Richmond SPCA, Maddie's Fund applicants who provide detailed information and are subjected to multiple layers of scrutiny), publish credible numbers, it won't be long until shelter people and lay folks alike have enough information to know when something looks fishy. When it does, ask for an explanation. Shelters have an obligation and a responsibility to fully disclose facts and figures about their operations. Animal people have a responsibility to ask for it.

Let's get back to the deaths per 1,000 issue. This is something Maddie's Fund has chewed on extensively and, in fact, I'll be publishing a detailed editorial on it in an upcoming issue of our Newsletter ( But let me lay out a few comments now.

Deaths per 1,000 is an important statistical measure. The data is relatively easy to collect, it can be applied in every community, and it is very objective. But it's only one measure to consider. Let me give you an example. New York City animal shelters kill 35,000 cats and dogs per year. New York's human population is 8 million. Therefore, deaths per 1,000 come to 4.375. This puts New York City 4th in the Animal People ranking of deaths per thousand. But does this really mean New York is doing an outstanding job of saving lives? Fourteen thousand healthy/adoptable animals and 6,600 treatable animals are still dying each year.

Deaths per 1,000 is relevant and helpful, but it doesn't provide the ultimate answer. It doesn't take into account the number of animals per household, a number that varies widely throughout the country. As is clear from the example above, the number of pets per household is generally much smaller in an urban area than in a suburban area, with the highest numbers generally in rural areas. This difference in demographics interferes with using deaths per thousand as a universal measuring stick.

Before any conclusions can be drawn about a community's life saving efforts, a host of factors need to be considered: impounds per thousand, adoptions per thousand (market share), deaths per thousand, shelter deaths/adoptions as percent of impounds, and total deaths. As I said, I'll go into all of these factors in greater detail in an upcoming Maddie's Fund editorial.

Tracking stats about why animals end up in shelters and developing programs to address

Question from Jen:

What about tracking the stats of why animals are ending up in shelters in the first place? It seems that if we paid more attention to this, we would know which type of program we need to do - a feral cat program, or a behavioral program for dogs, or spay days geared towards Pitbulls or labs. Yet no shelters in our area even keep track of these types of things and they say that they don't have the time or staff and even if they did, people would just lie about why they are surrendering them in the first place and you wouldn't get good data. Do you have any thoughts on this? Is there a good form to use or software programs that track this?

Response from Richard:

There are a variety of different approaches shelters can use to stop the killing of homeless pets. Every one of them is legitimate.

Shelters can try a front-end approach to the problem by attempting to prevent it from occurring, and focus their resources on spay and neuter. Shelters can react to the problem at hand and focus their resources on adoption programs. Or, shelters can pursue a back-end approach, determining the reasons animals are entering the shelter and using this information to design appropriate prevention programs (e.g. dog behavior).

In a perfect world, shelters would have all the resources they need to deal effectively with any and all sheltering problems. That's not the reality in most of the country today. Shelters have to make tough decisions on how to allocate their resources, and many of those decisions will be based on the organization's purpose and priorities.

Getting back to your question: I think most shelters have a pretty good sense of how many ferals or Pitbulls are coming into the system. Whether that information is important or not is a matter of how the issue is defined and prioritized. So, is it important to collect the kind of data you list? Absolutely... if the shelter can or wants to use it. Let me give you a hypothetical.

Let's assume a shelter has X amount of money in its budget for spay and neuter, and it's using all this money to alter all of its animals prior to placement. Then let's say this shelter determines that feral cats are a problem and they'd like to fix it with a TNR program. Should the shelter take money away from its current spay/neuter program to start a TNR program? Should it take money from another one of its programs to start TNR? Should it try to fundraise around TNR even though past efforts haven't achieved great results? This situation is difficult and the choices are challenging.

In the end, decisions are all about priorities and resource allocation.

If you can apply the data you gather, then get it. But if you can't or don't want to do anything with the information, then it's of limited value and you might be better off focusing your efforts elsewhere.
Kindness to animals builds a better world for all of us.