Do dogs deter burglars?

The academic source is Kuhns and colleagues' 2013 UNC Charlotte study, "Understanding Decisions to Burglarize from the Offender's Perspective" — a survey of 422 incarcerated burglars, funded by AIREF, an alarm-industry research foundation (a bias worth noting, though dogs are not the funder's product). When burglars described how they selected targets, the top factors were the proximity of other people, escape routes, and signs of security — and inside that last bundle, dogs inside the home sat alongside alarm signs and outdoor cameras as a leading reason to move on.

The journalistic source is KGW-TV's 2017 project, in which reporter Kyle Iboshi surveyed 86 incarcerated burglars in Oregon. It is not peer-reviewed, but its findings converge: the strongest avoid signals were a car in the driveway, sound inside the house, and big, loud dogs. The same inmates reported that essentially all of them knocked or rang the doorbell first to test occupancy — which matters for dogs, as we will see.

Two different populations, two different methods, two different funders, one shared conclusion. In a literature this thin, that kind of convergence is rare, and it is the main reason dog deterrence deserves more credence than most of what circulates about signs and stickers.

Watch what a dog actually does when a stranger approaches. It detects early — hearing and smell put it on alert before any human in the house knows anything is happening. It announces instantly: barking is noise, attention, and alarm rolled into one, audible to the intruder, the household, and the neighbors at the same moment. It positions, moving to the door or window nearest the threat. And it escalates — the longer the stranger stays, the louder and more agitated the response gets. There is no window where probing it produces silence.

Notice what is missing from that list: biting. The deterrent value is not primarily the prospect of injury — it is the certainty of being detected and confronted, loudly and immediately, with attention drawn from exactly the people a burglar least wants involved. Recall one of the UNC study's top target-selection factors: the proximity of other people. A barking dog manufactures that factor on demand. It converts a quiet empty house into a loud, watched one.

The UNC data has a second finding that rhymes with this. Over half of the surveyed burglars said that discovering an alarm mid-burglary would make them abort. Real-time confrontation interrupts crimes in progress — and a dog is real-time confrontation on four legs, running before any human decision gets made.

Now the caveats, because this literature earns fewer decimal points than it usually gets. Both surveys are self-reports from burglars who were caught — the unsuccessful ones, by definition. The UNC study is industry-funded. The KGW sample is 86 people in one state, gathered by a newsroom. No randomized study of dog ownership and burglary exists, and probably never will. What we have is two consistent self-report surveys, not a measured effect size.

The qualifier in the Oregon data also matters: big and loud. A small, quiet dog produces little of the mechanism — less noise, less threat, less manufactured occupancy. And dogs can be defeated: distracted with food, soothed by calm familiarity, shut behind an interior door, or simply risked by someone determined enough. The surveys show dogs shift the odds meaningfully. They do not show dogs make a house untouchable.

A dog's mechanism runs entirely at the property, which is also its ceiling. While you are away, it cannot tell you anything. It records nothing, describes no one, and produces no evidence a detective or an insurer can use. FBI data for 2024 shows daytime residential burglaries (216,601) outnumbering nighttime ones (174,053) — burglars work your work hours — so the most common scenario is a dog handling an intruder alone, with you learning about it from a mess hours later rather than a message in the moment.

The other scenario is worse. Bureau of Justice Statistics data found a household member was present in roughly 28% of burglaries. A confrontation between an intruder and your dog, with your family home, is not a plan — it is a gamble with stakes on both ends of the leash. Dogs also sleep, age, get sick, and travel with you, and the mechanism goes wherever they go.

No — get a dog because you want a dog. It is a ten-to-fifteen-year commitment of care, cost, training, and attention, and an animal acquired primarily as equipment tends to be shortchanged on all four. The deterrence is a real side effect of a well-kept dog, not a reason to acquire one.

But the surveys leave behind something useful either way: a specification. The thing that deters is not fur or teeth — it is an on-site responder that detects early, confronts immediately, and escalates for as long as the intruder stays. Write the mechanism down like that and an obvious question follows: can it be automated?

That is precisely what an Automatic Property Defense System is engineered to do — reproduce the detect-confront-escalate loop without asking an animal to hold the line. Monitoring products watch and notify; the dog's whole lesson is that deterrence lives in what responds at the scene, in the moment.

PerimeterEye runs the loop in seconds. It verifies a person on the property, then a motorized searchlight locks onto them and physically follows them — 355 degrees of rotation, 90 of tilt, the effect homeowners compare to a police-helicopter beam. Then it speaks: live, targeted words describing the specific person in frame — "You, in the dark hoodie — you are not authorized to be on this property" — not a recording on a loop. If they stay, a 130 dB siren sounds and priority alerts with live video go to your phone, where you can watch, talk live, or take manual control. Detection, confrontation, escalation — the dog's mechanism, minus the nap schedule.

It even has the good dog's other trait: it knows the household. Family arrivals and the mail carrier never trigger deterrence, the same way your dog does not bark at your own kids. No security system can guarantee prevention of every incident — but the mechanism both burglar surveys point to can now run on watts instead of kibble.