On April 25th, a man in Fortuna, California sent a shelter worker a photo of a dog named Zora on a leash, with two words: “Zora adopted.” Days later, Zora’s body was recovered from a burial site behind Miranda’s Rescue. She had never been adopted. She had been shot.

Miranda’s Rescue was a registered nonprofit that called itself a no-kill sanctuary. Shelters across California sent it their hardest dogs, and in many cases paid it to take them, between $400 and $1,000 per dog. Investigators have since recovered the remains of at least 138 dogs from the property, with more than 700 still unaccounted for.

This episode isn’t really about one man. It’s about how a system full of people who believed they were saving dogs delivered animals and money to that property for years, with no reliable way to know what happened next. Penny walks through the four forces that made it possible: a law that pressures shelters to release even dangerous dogs to any willing rescue, a “live release” statistic that counts a transfer as a save the moment the dog leaves, a payment structure that rewards intake instead of outcomes, and a reporting system built to confirm that a dog left but never to confirm where it landed.

Then she turns to what might prevent a tragedy like this from happening again. Colorado has required licensed shelters and rescues, including foster-based ones, to file a public intake-and-outcome ledger every year for more than two decades, and Penny explains both why that ledger is the model and where even it falls short. She closes with two things any shelter can do right now, without waiting for a legislature, and the questions every advocate should ask before recommending a rescue or sanctuary to anyone.

In this episode, you’ll learn:

• Why paying a rescue at intake, rather than for outcomes, creates an incentive that can be quietly exploited
• How a transfer counted as a “live release” can conceal what actually happened to a dog
• Why vetting a rescue at the front end may not be enough, and what to look at instead
• How Colorado’s public intake-and-outcome ledger works, and where its reporting timeline leaves a dangerous gap
• Two things any shelter can do right now to make deception risky: a stronger transfer agreement and a simple spot check
• The questions to ask before you recommend a rescue or sanctuary, or send your money

Key Takeaway

A dog leaving a shelter is an event, not an outcome. Getting an animal out the door is only rescue if someone has to report what happened later.

Episode Highlights

00:01 — The text that said “Zora adopted,” and the body recovered days later
00:56 — Shelters paid $400 to $1,000 a dog, with no way to verify outcomes
02:32 — How Jennifer Raymond and Sabrina Woods noticed the pattern and started digging
03:19 — Hard-to-place dogs reported adopted almost overnight, with no visible effort to place them
04:59 — Manufactured happy endings: false adoption photos sent to stop the questions
06:15 — Follow the money: how pledges and paid transfers can move a dog without securing a safe destination
09:11 — Why a “live release” counts a transfer as a save the moment the dog leaves
12:04 — Why the hardest dogs were the easiest to make disappear
16:44 — Why front-end vetting can pass an operation like this 
17:13 — Colorado’s public intake-and-outcome ledger, and where its reporting timeline falls short
22:31 — Two things a shelter can do now: a tighter transfer agreement and a simple spot check
28:19 — Be the Change: find out what your state requires of rescues, and ask for public reporting

Resources Mentioned

• Colorado’s public shelter and rescue statistics (PACFA)
• Pennsylvania kennel inspection search (Bureau of Dog Law Enforcement)
• The Four C’s of Legislative Advocacy for Animals (free audio series)

Related Episodes:

How to Help Animal Shelters on Social Media: Do’s and Don’ts from Shelter Staff

Why Animal Advocates Need a 501(c)(4): Allie Taylor of Voters for Animal Rights

Transcript 

On April 25th, a man in Fortuna, California sent a text to a shelter worker in Oakland. Attached was a photograph of a dog named Zora on a leash with a message that said, Zora adopted. Days later, two women dug into a burial site behind Miranda’s Rescue and recovered the bodies of eight dogs. Zora was one of them.

She had not been adopted. According to a sheriff’s affidavit, the man who sent the text with the adoption photo later admitted that he had shot her. Investigators say he also admitted sending false adoption messages about other dogs he’d killed, explaining that he did it to protect feelings.

 Now, Miranda’s Rescue was a registered nonprofit that described itself as a no-kill rescue and sanctuary. Shelters across California sent it some of their most difficult dogs, and in many cases paid the rescue between $400 and $1,000 for each dog it accepted. Investigators have since recovered the remains of at least 138 dogs, along with hundreds of additional bones from the property.

The question I want to explore today isn’t just how one person could do something this horrific. It’s how a system filled with people who believed they were saving dogs delivered animals and money to that property for years without any reliable way to know what happened next.

Welcome to The Animal Advocate, where we arm animal lovers with the information and inspiration you need to become effective advocates. I’m your host, Penny Ellison, and I’ve taught animal law and advocacy at the University of Pennsylvania since 2006. If you’ve ever thought, someone should do something about that, I’m here to guide you on your journey to being that someone. You can find us on the web at animaladvocacyacademy.com. And that’s where you’ll find show notes and resources, and you can send us your comments on episodes and ideas for topics you’d like to hear on future shows. So, on to today’s topic.

The discovery of 138 dogs buried at Miranda’s Rescue did not come out of nowhere. For years, animal advocates had heard concerns that dogs sent to Miranda’s seemed to disappear. Jennifer Raymond became so suspicious that in 2023, she bought the property next door so she could observe what was happening behind the rescue’s fences. From her home, she watched dogs arrive and then vanish from the kennels. When she asked about particular animals, she said she was often told they’d been adopted or were unavailable. At the same time, a Solano County shelter volunteer named Sabrina Woods was also noticing something else that didn’t add up. Dogs that shelters had struggled to place, including difficult behavioral cases, were reportedly being adopted soon after arriving at Miranda’s Rescue, even though there was little visible effort to advertise them.

Raymond and Woods began seeking shelter records showing how many dogs had been sent to the rescue, how much money had been paid, and what outcomes had been reported. The records were staggering. Miranda’s Rescue had taken in more than 900 dogs just since the beginning of 2025, yet investigators could initially confirm only about 100 adoptions. Hundreds of dogs, and eventually more than 700 by the sheriff’s count, remained unaccounted for. Then Raymond saw a fresh mound of dirt near the edge of the property. On April 12th, she entered the property and dug into it. She uncovered the body of a dog.

She and another advocate, Jenna Moore, then positioned motion-activated cameras toward that field. Two weeks later, the cameras reportedly captured activity around another burial site. Raymond and Moore returned with shovels, headlamps, and a microchip scanner. They recovered eight recently buried dogs. Six had microchips. One was Zora, the dog whose adoption photo had gone out the day before. She was never adopted. She was in the ground. While a shelter worker in Oakland looked at that photo and celebrated her happy ending.

This case goes way beyond any lack of transparency. This was a rescue manufacturing false happy endings, adoption stories built to stop anyone from asking any questions. A dog can leave a shelter alive, be counted as a successful transfer, and then disappear almost completely from view. To understand how that happens, you have to follow the money. And the money starts with people who love animals.

When a dog is facing euthanasia in a crowded shelter, money can seem like the answer. People pledge money for the rescue that’ll save the dog, so it can be used for transportation, medical treatment, boarding, training, whatever’s necessary. And rescue costs a lot of money, so sometimes that money makes a life-saving placement possible.

But money can also move a dog without securing a safe destination. A funded transfer is not necessarily a rescue. The critical question is not simply whether enough money was raised to get the dog out of the shelter. It’s where the dog’s going, whether the organization has the capacity to provide what it’s promising, and how anyone will know what ultimately happened. That distinction is at the heart of this case and the problem that it illustrates. In Oakland alone, the nonprofit affiliated with the city shelter reportedly paid Miranda’s Rescue approximately $376,000 over several years. Much of that money came from donors who believed they were helping hard dogs leave the shelter and start a new life.

In the ordinary shelter-rescue relationship, the rescue agrees to take responsibility for a dog because it believes that it has a foster home, adopter, a treatment plan, or other viable placement like their own kennel. Often, and usually I think, no money changes hands. In some systems, and increasingly recently with shelters around the country desperate to create space, an overwhelmed shelter or its donors or funders might pay a rescue to accept animals that they’ve been unable to place. Now, that arrangement is not inherently improper. It may pay for all those expenses of accepting a medically or behaviorally difficult dog, but it creates an incentive that has to be recognized. If the rescue is paid when the dog arrives, the financial reward is tied to intake, not outcome. Once the payment has been made, every additional day of food, housing, vet care, and staffing costs the rescue more money. A responsible rescue accepts those costs because caring for and appropriately placing the dog is its mission.

 An unscrupulous operator can make more money by taking in more dogs and spending less on each one. That doesn’t mean, of course, that every paid transfer is suspicious. It means a system that pays at intake without requiring any credible outcome records is relying almost entirely on trust. At Miranda’s Rescue, that trust appears to have been exploited in the most brutal way possible.

How does a dog go from a municipal shelter surrounded by staff and volunteers to a hidden grave behind a rescue? There’s no single cause. Several forces pushed in the same direction, and each one looks understandable when you view it alone.

The first was the law. California’s Hayden Act requires a shelter to release an animal scheduled for euthanasia to a qualifying nonprofit rescue that requests the animal before the euthanasia occurs. Now, the statute uses the word shall— shall transfer— and a court has held that that duty is mandatory. The legislature considered limiting that requirement to animals the shelter considered adoptable or treatable, but it didn’t.

Now, the law serves an important purpose. It prevents a shelter from killing an animal when a legitimate rescue is ready and able to take responsibility for that animal. But that mandate creates pressure when the dog has a bite history or presents risks that the shelter does not believe it can safely manage. A rescue may appear and say it specializes in exactly those cases. So, the law favors release. The shelter’s desire to avoid euthanasia favors release. The public pressure to save the dog favors release. The rescue becomes the answer to a problem that no one wants to deal with.

Now, the second force at play is the way shelters measure success. The live release rate is the percentage of animals who leave a shelter alive through adoption, return to an owner, or transfer to another organization. That number becomes very important because it matters to the public, to funders, to government officials, and to shelter leadership. Also, it’s an understandable measure. People want to know whether animals entering a shelter are leaving alive. But a transfer is counted as a live outcome when the dog leaves the shelter.

That does not mean the shelter is claiming to know that that dog will live safely for the rest of its natural life. It means the shelter’s responsibility has been transferred to another organization. Open-intake shelters could not function if they were expected to follow every animal through every later foster home, adoption, return, veterinary decision, change in circumstance. The problem arises when a transfer is publicly understood as proof that the dog was saved while the receiving organization’s outcomes remain invisible.

The third force at play was the payment structure. Miranda’s Rescue was paid, as we said, about $400 for an ordinary dog and as much as $1,000 for a dog with a bite history. The more difficult the dog, the more money the rescue got. Those dogs were also the easiest to make disappear. And I’ll tell you why. A friendly, adoptable dog probably has people watching— volunteers, potential adopters, social media followers tracking his story. A dog with a bite history often has no one. And when word finally comes that a dog like that found a home, the people who’d worried about him are usually so relieved, so thrilled to hear that he made it, that the last thing they’d think to do is to ask a hard follow-up question. A live outcome for the hardest dog feels like a gift. You don’t interrogate a gift. Miranda’s Rescue was reportedly paid the most for precisely the dogs least likely to have someone looking for them.

And then there was the fourth force. Almost everything depended on the rescue truthfully reporting what happened. A message arrived saying, Zora adopted. It included a photograph of Zora on a leash. The message was apparently designed to answer the question before anyone asked another one. The deception worked because the system was built to confirm transfers more carefully than outcomes. Shelters recorded the dogs leaving. Donors funded that rescue. The receiving organization supplied the ending. Zora adopted. So, the shelter had that live release for its statistics. The donor was thrilled that their donation helped Zora get a happy ending. The rescue got its money. The only participant in the transaction who didn’t get the outcome they needed was Zora.

There’s another argument surrounding this case that I want to address directly. Some people have responded that these were dangerous dogs and that euthanasia may have been the appropriate outcome for them, implying, you know, something like no harm, no foul. Some of the dogs may in fact have presented serious and unmanageable safety risks, and behavioral euthanasia is sometimes necessary. It wouldn’t be honest to claim that every dog can be safely placed.

But that does not excuse what happened here. When a shelter determines that a dog is too dangerous to release, that decision should be made openly based on the individual dog’s behavior and other available evidence. It should be documented, counted, and owned by the organization responsible for making it.

What appears to have happened here was the opposite. The decision was laundered. The dog left the shelter as a recorded live release. A rescue accepted money based on the representation that it would assume responsibility for the dog. Donors believed they were supporting rescue rather than euthanasia. Then, according to the allegations, some dogs were neglected, killed in secret, and buried while false adoption reports were sent to the people who might otherwise have asked what happened. A transparent euthanasia decision can be reviewed, questioned, counted, and criticized. A transfer followed by a secret killing converts a recorded save into an invisible death. The problem isn’t simply that some difficult dogs may ultimately have to be euthanized. That happens. The problem is that those deaths were concealed. From the donors, from the public, and even from the shelter that sent the dog.

So, what should shelters have done differently?

 It’s tempting to say they should have investigated the rescue more thoroughly before transferring animals. They should have confirmed its nonprofit status, spoken with other shelters, checked references, toured the property, and verified that it had been providing good veterinary care to the animals there. Those are all reasonable precautions. But the uncomfortable lesson from this case is that Miranda’s Rescue looks as if it probably would have passed those usual due diligence tests. It wasn’t some newly created organization operating from an unknown address. It held government contracts. It had a large property. Shelter representatives toured it and saw living dogs. Other organizations gave them favorable references. The operation had enough external legitimacy to reassure people who made reasonable efforts to evaluate it. That’s why the answer can’t just be to screen harder at the front end. Nearly all the scrutiny in this system pointed at the front end. Almost none of it pointed at outcomes.

 So what would catch this kind of thing? Not more front-door screening. What catches it is finding some way to monitor what happened after the rescue took the dog. And the good news is, we don’t have to invent it. At least one state has been doing it for the last 26 years.

Colorado licenses shelters and rescues, both, including the foster-based ones with no building. If you handle more than 24 animals in a year, you’re licensed, or you’re supposed to be. And once a year, you file a report on every animal that came through your hands. The state posts it online, and anyone can pull it up. I went and looked at it, and I want to tell you what’s on it because it’s simple enough to explain in about 30 seconds.

For each rescue, the state publishes how many dogs you had on January 1st, how many you took in. Where they came from: stray, owner surrender, transferred from another Colorado group, or transferred from out of state. Then, how each of those animals left you: adopted, returned to owner, transferred out, died, euthanized, missing. And finally, how many you still had on December 31st. Start of the year, in, out, end of the year. That’s a ledger, and a ledger has to balance. If a rescue reports 900 dogs in, 100 adopted, 0 deaths, 0 euthanasias, and 20 animals on hand at the end of the year, that arithmetic doesn’t add up, and anybody can see it. You don’t need an investigator, you just need a calculator. Colorado even publishes a column for animals that went missing.

Now, here’s what’s reported in my own state, Pennsylvania. Pennsylvania licenses rescues too, including foster-based ones, which is great and not necessarily common. We call them Rescue Network Kennels, and the foster homes register under the license. The state inspects them and puts the inspection reports online, free, searchable, for anyone. I pulled one at random, a boxer rescue up near Scranton. The report tells me the license class, which is generally how many dogs pass through the facility in one year, the inspection date, that the records were kept properly, and a table of dog counts. And that’s where it stops. There’s no column for adopted, died, euthanized. Pennsylvania tells you publicly how many dogs went into a rescue, but not about the outcomes.

 So picture what that report would’ve looked like for Miranda’s Rescue. A satisfactory on the records kept, and every dog buried in that field invisible in the public record. So, in Pennsylvania, the ask is almost embarrassingly small. The database exists. The rescue is already filing. The inspector is already visiting. The form already has a table with numbers on it. Just add columns to show the outcomes. That would be great.

 But I also want to be honest about the limits, including Colorado’s. It’s now July of 2026. The most recent report Colorado has published covers 2024. That’s a year and a half old. And depending on when you check, it can be closer to two and a half. Think about what that means for our story. Miranda’s took in more than 900 dogs starting in January of 2025. Under Colorado’s timeline, not one of those dogs would appear in a public report yet. The scheme ran and collapsed and got dug up entirely inside that reporting gap. So the ledger was right. But the clock is wrong. A rescue moving hundreds of dogs a year knows its numbers every month. It has to. That’s how you know whether you can say yes to the next one. Filing those numbers four times a year instead of just once, and posting those reports as soon as they arrive, should be achievable.

And one more caveat: Colorado doesn’t audit these numbers. The state says so right on the page. The numbers come from the facilities, and if you have questions, ask the facility. So, a determined liar, like Miranda, can file whatever he wants, and the arithmetic will balance, if he’s not stupid. But he has to file it every year under his license on a public page where the shelters who sent him dogs can read it. And a claim of 900 adoptions is 900 claims that somebody, somewhere, took a specific dog home. That’s a lot of people to invent. And if a shelter asks questions and there’s that many made-up outcomes, they’re going to figure it out. It’s not a foolproof system, but it’s a lot better than what we have now.

So that’s what a state can do. But we don’t have to wait for the legislature to act, thank God, and neither does a shelter transferring dogs to rescues. There are 2 things a shelter can do right now, and neither one requires following a dog for the rest of his life.

The first is to tighten the transfer agreement. It should say what the rescue can and can’t do with the animal. No euthanizing a dog just to make space. No handing the dog off to a third party you never approved. Because a shelter can decide not to send dogs to a particular rescue and still have its dogs end up there, routed through some other group acting as a middleman. And it should require the rescue to register the adopter on the microchip when the dog is placed, with the rescue staying on as backup so they’re still reachable if the dog is ever lost. That one line does two things: it gets the dog home if she gets out, and it means every adoption should be recorded somewhere other than the rescue’s own files.

The second thing shelters can do is a spot check. A shelter can’t track every dog, and it doesn’t have to. It only has to check enough of them to make lying dangerous. Pull a handful of dogs a high-volume partner reported as adopted and see whether the chip registrations ever moved. That’s free and it doesn’t take long. If they didn’t move, start asking. Who adopted her? If the dog is supposed to still be in sanctuary, can someone confirm she’s actually still there? A few calls, maybe, a quarter. One fabricated adoption is reason for a much closer look. A pattern of unchanged chips and adopters nobody can reach is reason to stop sending dogs. It just makes sense. It’s what you’d do with any organization you’re handing vulnerable animals to, and in some cases, paying to take them.

So some shelters saw this story and they’re already doing exactly this: rewriting their agreements, reviewing every rescue on their list, building in a system of verification. The point is not to make transfers so burdensome that shelters lose the partners they depend on. Open-intake shelters need those partnerships. A lot of dogs are alive because a rescue said yes when a shelter ran out of safe and humane options. But there’s a wide gulf between demanding that a shelter follow every animal forever and taking a paid transfer operation’s word for it every time, forever. A contract, honest records, a chip registered to the right person, and a few calls are not unreasonable burdens. A good rescue is already doing all of it, and they’ll be glad you asked. It only troubles the operation that can’t answer the question.

 And there’s also a lesson here for the people who share urgent rescue pleas online, and there are millions of them. When a dog lands on a euthanasia list, someone inevitably tags some nationally known sanctuary. You know the kind. They have a great website, millions of followers, videos of dramatic transformations. People repeat the name because they’ve seen the content so often that the organization feels familiar. But knowing the brand is not knowing the rescue. A polished social media presence can show real and important work. It can also show only the part of the operation that someone chose to put in front of a camera.

And be especially careful with the ones advertising themselves as a sanctuary. Sanctuary is the most expensive promise in this field. Long-term care for the animals nobody else will take, which for an old or sick or dangerous dog means trained staff, vet treatment, medication, housing for years. It’s also the promise least likely to be checked, because the dogs are supposed to still be there. There’s no new outcome to check, so nobody goes looking.

Before you recommend that a dog be sent somewhere, especially a sanctuary, or encourage strangers to donate, ask yourself whether you could answer the basic questions about that organization. How many animals are there? How are they housed? How many people are caring for them? Are they receiving vet care? How many were adopted or euthanized or died in the last year? Those aren’t hostile questions. They’re questions that distinguish a rescue operation from a compelling story about a rescue operation.

And notice, the question isn’t whether the organization could answer those questions. It’s whether you can. If you can’t, you don’t know the rescue. You know it’s marketing. You would never recommend a breeder just because the breeder posted beautiful photographs of healthy puppies. Rescue organizations shouldn’t be exempt from the same basic skepticism because their marketing appeals to compassion instead of consumer desire. Because operations like Miranda’s Rescue don’t build reputations alone. Their reputations get built through social media posts and recommendations repeated by people who know the organization only from a distance. Every one of them was trying to help, and at least 138 dogs paid for it.

And now, our Be the Change segment. Here’s your one Be the Change action for this week, and it doesn’t take a law degree or a single dollar. Find out what your state requires of the private organizations that take animals. The rescues, the sanctuaries, the foster networks, the ones with a building and the ones without. Does your state license them at all? And if it does, does anyone have to say what happened to the animals? Search your state’s name and animal rescue license or kennel license, or call your state’s Department of Agriculture and ask who licenses rescues and whether the reports are public. It’s a 10-minute phone call. Then you’ll know which conversation you’re having.

If your state licenses rescues and publishes something, you’re in Pennsylvania’s position. You have infrastructure, and your ask is small and specific: put the outcomes on the report. If your state licenses rescues but publishes nothing, your ask is to make public what’s already being collected. And if your state doesn’t license rescues at all, that’s your starting line. It’s a bigger lift, but you’re not asking for anything crazy. You’re asking for what Colorado has been doing for more than 25 years.

 The most important lesson from this case is not that shelters should stop transferring difficult dogs or that the public should stop supporting rescues. Both would cost a lot of lives. The lesson is that transfer can’t be the end of the story.

A shelter may never be able to know what happens to every animal for the rest of its life, but an organization entrusted with animals should keep credible records. The agencies that send dogs and money should check enough of those records to make deception risky. And the rest of us should look past the photographs before we help an organization grow. Because a dog leaving a shelter is an event. It is not, by itself, an outcome. “Zora adopted”  was an outcome on paper. It supplied the ending everyone wanted. The shelter worker could believe she was safe, the rescue could close the file, and nobody had to ask a single question.

We owe the dogs more than that. We have to follow the dog far enough to know that the rescue was real.

That’s it for today. The Animal Advocate is brought to you by the Animal Advocacy Academy. If today’s episode got you thinking about what you can do to change things for animals, I want to make sure you know about the four Cs of legislative advocacy for animals. It’s a free audio series I put together that gives you a real framework for making legislation happen. You can find it at animaladvocacyacademy.com/fourcs. and you can find all our episodes and show notes at animal advocacyacademy.com. Subscribe so you don’t miss an episode, and if you know another animal advocate who should be listening, send this one their way. Questions on today’s topic or anything else, just email us at podcast@animaladvocacyacademy.com.  We’ll write back, or we’ll feature your question in a future episode.

Remember, compassion is great, but compassionate action is infinitely better. Until next week, Live with Compassion.