Bias in the perception of poo pile concentration in dog parks

(6-minute read)

Non-dog people like to complain about how much dog poo they encounter all over creation—and how they feel it reflects dog people’s sloppy, or even willful, disregard of public health and aesthetics. As an ardent dog person, I believe that compliance rates of dog poo pickup are generally quite high, and I just do not see dog poo littering the earth. Which perspective is more accurate?

The dog poo perception problem is an example of cognitive bias: a behavioral economics concept developed by, among others, Daniel Kahneman (Nobel 2002) followed by Richard Thaler (Nobel widely agreed to be forthcoming), and lately popularized by Malcolm Gladwell (Tipping Point, Blink) and Steven Levitt (Freakonomics). We humans are “predictably irrational” thinkers, as a result of laziness, self-interest, emotional lability, fear, and social pressure. Perception of dog poo concentration is just that: perception. And human perception is often quite biased and thereby inconsistent with observable reality.

A poo by any other animal would smell . . .

Non-dog people are emotionally motivated to have beliefs about high incidence of dog poo, and then to operate with confirmation bias regarding their preconceptions. Take my non-dog ex-cousins in Minneapolis, for example. An ill-fated Thanksgiving holiday visiting their home involved high-stepping around hundreds of gooey, green-and-white piles of poo. One unintended consequence of arriviste home-buying in pond-laced designer neighborhoods turns out to be . . . geese! Geese were all over the street, on the front walk, and inside their expansive, cloistered yard. 

The cousins, having no control over these honking symbols of suburban wealth, were blind to the pervasive goose poo problem. In contrast, they became outraged over one single pile of dog poo from the neighbor’s small terrier squeezing under the fence. The dog, being theoretically controllable, was infringing on their high-priced entitlement to privacy. Diplomatic allusions to the hypocrisy of this attitude were dismissed flippantly; so my then-husband and I were left to commiserate with one another wordlessly through a days-long monologue by those insufferable cousins about the alarming dog poo problem in society at large. Given that they expect to see poo everywhere, they pay disproportionate attention when they run across it, and consequently perceive it to be everywhere.

Moral consequences

Confirmation bias is also the insidious culprit in climate change denial. The same type of selective information gathering and emotionally-motivated reasoning that enables a rancorous minority to perceive a catastrophic dog poo problem, enables a (probably quite congruent) minority to deny a catastrophic greenhouse gas problem. They cherry-pick evidence that conforms to their pre-established opinion (opinion which arises from personal values, not from logical reasoning), no matter how specious. When presented with contrarian evidence—no matter how voluminous or solid—, they typically react by doubling down on that same pre-established opinion. Climate change deniers are practicing a form of religion: they require no solid evidence to support their faith, and no new facts would ever change their minds.  

In comparison to anthropogenic climate change, the dog poo debate may seem mundane. But it, too, is a serious problem. The treatment of animals (including accuracy in assessing their nuisance level in terms of leftover poo) reflects the morality of a society. Sloppy reasoning, unchecked on a small scale and then applied on a grander scale, can do great harm.

How much poo is actually out there?

How much poo is actually lingering out there in our otherwise safe, verdant parks? Let’s do a back-of-the-envelope calculation to get a sense of expected poo concentration in city dog parks:

• Consider park acreage and dog population of a given city. (I have used my old hometown of Portland, Oregon, as the example.)
• Make assumptions about park visits and poo-deposition habits.
• Make an assumption about the poo pick-up rate. Consider that there are legitimate reasons that poo is sometimes not picked up: human was chatting with another human and didn’t see it happen, poo deposited at night can be impossible to find, responsible humans with bags periodically run out of bags.
• Make an assumption about poo half-life (poo accumulates as it’s abandoned, but degrades and disappears over time)

Given the default assumptions shown in the calculator here, we might expect to come across one pile of abandoned dog poo about every 1,000 square feet in a city park. In fact, that calculated poo concentration happens to be generally consistent with observed reality in my neighborhood parks.      

That’s a lot of poo!

Frolicking dogs and their poo-prints are concentrated in certain sections of parks, rather than evenly distributed across the entire acreage. Therefore, poo will be clustered into hot zones. A grumpy non-dog person meandering through a park might run into such an area, with poo littered around far more densely than 1 per 1,000 square feet. Such clusters are vividly memorable when encountered. Despite the vast poo-free acreage, it is the poo clusters to which the irrational human mind pays attention, and which is then perceived as general reality.  

My calculator does not contemplate non-compliance due to willful negligence (though if you want to include that, just lump it into the assumption for unintentional non-compliance). Indeed, we do not need to assume any willful negligence at all, to explain the proliferation of poo. Even with very low rates of (unintentional) non-compliance, there are just so many dogs that it still adds up to a ton of poo.

So, we can conclude that it is a myth that dog poo in city parks is due to willfully irresponsible dog owners. In fact, according to our model, rare and innocent lapses among responsible dog owners are sufficient to explain a heck of a lot of poo.

But . . . is that really a lot of poo?

How high of a compliance rate would it take to make poo a vanishingly rare sighting? The answer is that it would have to be impossibly high. Even with a super-high compliance rate of, say, 95%, the issue is that there are still just a lot of dogs out there. 

Moreover, regardless of whether higher compliance would make a noticeable difference in poo concentrations on the ground, aiming for higher compliance is not terribly realistic. (Of course, if, in some localities, there indeed were a verifiable problem of willful non-compliance, that obviously should be addressed.) The 12% non-compliance rate assumed in our default model (i.e., 88% compliance, a level that produces a projected poo concentration matching observed actual poo concentration) would already be considered an admirably low rate of leakage in most regulatory systems. Absolute adherence to any rule is not a reasonable expectation.

For example, the compliance rate for handwashing among patient care workers in U.S. hospitals averages just 40%. Hospitals often set an aspirational target of 90% handwashing (note: not 100%!) in compliance improvement programs, not even confident they meet it despite the health consequences at stake.

Some other compliance/adherence rates in the U.S. to consider, for context:

• 48% of adults who have been prescribed medication for chronic illness comply with the prescribed drug, dosage, and timing
• 58% of citizens vote in general elections
• 65% of adults maintain their body weight below the level of obesity (though many are still medically overweight, just not obese)
• 86% of drivers have auto insurance (as low as 70% in some states), even though it is mandatory and enforced with hefty fines or even imprisonment
• 91% of households who must file a federal tax return actually do so

So, how about we throw the poor dogs a bone and give their owners a break? Upon analysis, it seems that poo pick-up compliance actually may be pretty darn good!

To what extent could my own assumptions, calculations, and conclusions themselves affected by cognitive and motivational bias? Use the interactive calculator below to explore the issue for yourself by adjusting assumptions in the poo concentration prediction model: