Each year we get a call from a media outlet doing a story on the sales of bunnies around Easter and the inevitable subsequent flooding of animal shelters with rabbits. My response has generally been, “Holiday marketing of pets isn’t a great idea and people should think carefully about any pet, however, we aren’t overrun with rabbits.” The media outlet would usually end the quote at “pet”.
This year I went a step beyond doing my usual hard census count of the number of rabbits entering our shelters in the past year and reminding myself how few rabbits actual come us. I compared that intake amount to the rates of household ownership of rabbits, cats and dogs and, somewhat to my surprise, we actually do have a tsunami of rabbits entering our shelters each year. At least statistically. In one category. In one county.
My first quick math taking only the relative intake numbers based on total numbers of households owning rabbits, as opposed to actual number of owned animals in those households, gave me a surprising number. It looked like there was between a one 150% and 300% greater likelihood of a pet rabbit being surrendered to a shelter in Berks or Lancaster Counties, depending on county and whether compared to cats or dogs. I was surprised, to say the least. It nagged at me for a day and I went back to the numbers to make sure I controlled for actual pet population in the households, not just households.
Based on AVMA pet population statistics, there should be 21 pet dogs and 23 pet cats for every one pet rabbit on average. That means that if cats, dogs and rabbits all entered shelters at the same rates, we should be able to determine the number of cats and dogs entering shelters by simply multiplying the number of rabbits received in a year by 21 or 23. If that multiple is close to the actual intake, the intake rate would be about the same for all. In the actual intake of cats and dogs is higher than the multiple, it means that rabbits enter at lower rates based on population (and a lower multiple would mean the converse).
Based on a rabbit intake number of 39 over 365 days in Lancaster, we showed an intake rate that was 56% over what we would expect when compared to dogs! A veritable tidal wave of bunnies! But compared to cat intake in Lancaster, that number dropped to about 2% greater likelihood for rabbit surrender. A ripple of bunnies. When I used our Berks intake statistics it shifted even further. Rabbits were half as likely to enter our shelter there than both cats and dogs. What gives?
I think it serves as a great reminder that we can use math to prove just about anything if we select the right inputs. My first pass at this used a reasonable measure- households owning rabbit, cats and dogs- and it gave me arguably defensible but simply wrong output numbers. Sound the alarms, rabbits at Easter end up in shelters in droves! But when I accounted for the numbers of rabbits actually owned, the numbers say exactly the opposite.
This is reminiscent of the City of Reading’s aggressive breed ordinance, which triggered when a certain number of bites occurred by a single breed. That breed was always pit bulls. The only problem was the number of bites was actually equal to or lower than the actual percentage of pit owned in the city. There was not a pit bull bite epidemic, there were just a lot of pit bulls. We found the same thing when we did internal research on the “epidemic” of big, black, dogs not being adopted in shelters. Nope, we found they faced the same odds as anything else when we controlled the numbers for intake by breed and size. Floods of animals related to those appearing in Disney movies? Nope. Black cat sacrifice at Halloween? Nope.
Sometimes the numbers just aren’t there, no matter how you count them. Sometimes we focus on a count and a number and decide that number means more than it actually does. Other times we can actually find a statistical tsunami, like the- gasp- 56% greater likelihood of rabbits to end up in a shelter compared to dogs in Lancaster County. But sometimes an earthquake creates a tsunami that crashes onto shore as a six inch wave. It’s real, but not a concern. When we are talking about 39 rabbits and a 100% rabbit adoption rate at our shelters, I’ll put my concern elsewhere.
This is also a reminder that just doing some quick math can get you in trouble. My first round was wrong. My second round is adequate but won’t be getting published in Nature any time soon. It doesn’t account for so many things, such as actual ownership in our region, alternative relinquishment outlets, that the source statistics I used were correct, that my methodology and math are even right (I got out of my astronomy career track for a reason, after all). And sometimes really big and really small numbers can do suspicious things to calculations and can radically swing from year to year. Beware “simple” math and certainty based on one man’s numbers.
But it also reminds us that we cling to old narratives. Easter bunnies; big, black dogs; and Disney Dalmatians. As comfy as an old sweater and probably also overdue for the trash bin.