SAN DIEGO, CA— On June 19, 2025, a New York attorney filed a federal lawsuit in the Eastern District of New York. She didn’t file it for a spouse, a child, or an aging parent. She filed it “on behalf of FINNEGAN MARY REYNOLDS,” described in the complaint as an “eight year-old American-born golden retriever.”
The complaint calls the question “novel but urgent”: whether “domestic companion animals… may be recognized as non-human dependents under U.S. federal tax law,” and whether excluding them violates equal protection and the Fifth Amendment’s Takings Clause.
It’s the kind of story built for the internet: cute dog, taxes, outrage, a dash of constitutional language, and the implied accusation that the government doesn’t “recognize” what anyone who has ever paid a vet bill already knows—pets are financially dependent.
But the seductive framing (“my dog is dependent, therefore dependent status”) collapses the actual meaning of dependent in the tax code. And if you do collapse that meaning, you don’t just create a dog-sized exception. You blow a barn-door-sized hole through the concept of limits.
I say that as a dog owner. I also own livestock. I know exactly how quickly “pet” turns into “dependent” once you define the term by who eats out of your wallet.
What the lawsuit actually argues (and what it reveals)
The complaint’s core claim is straightforward: Reynolds says dog owners’ financial contributions “meet or exceed the definitional criteria of ‘support’ under 26 U.S.C. Section 152,” but the IRS denies “tax recognition of that dependency.”
It goes further:
It asserts dogs should be recognized as “quasi-citizens entitled to limited civil recognition,” including dependency status.
It alleges Finnegan is, “for all intents and purposes… like a daughter, and is definitely a ‘dependent’.”
It points to the fact that certain service animals can be treated as deductible medical expenses (via IRS Publication 502) and argues that shows a “dog” isn’t inherently disqualifying for tax preference.
That last point is rhetorically clever—because it borrows legitimacy from a real, existing rule (limited deductibility for service-animal-related medical costs) and tries to convert it into a sweeping entitlement (“therefore all dogs should be dependents”).
But that’s a category error. The tax code already contains narrow animal-related rules because it recognizes specific human policy objectives (health accommodations, business expenses, charitable contributions). That is not the same thing as turning “dependent” into “anything alive I pay for.”
The law’s plain-language problem: “Dependent” is a defined term, not a vibe
Here is the spine of the entire issue. The Internal Revenue Code defines “dependent” like this:
That is not poetic. It’s a gate.
And when the statute then defines “qualifying child” and “qualifying relative,” it repeatedly uses “individual” and builds tests around human relationships (child, sibling, foster child, spouse filing status, custody rules), human residency/citizenship status, and human identification systems (taxpayer identification numbers).
The IRS puts it even more bluntly in its public guidance:
A golden retriever does not fit inside those words without rewriting them.
Which is exactly why, in December 2025—months after the complaint—an EDNY magistrate judge granted a stay of discovery while the government prepares a motion to dismiss, and wrote the part that should have ended the debate:
“the Tax Code simply does not allow for animals to be claimed as tax dependents.”
The order also notes the government’s stated rationale: requiring dependents to be human is “rationally related to the legitimate government interest in preventing fraud and abuse in the tax system.”
That’s the legal reality. Now for the policy reality.
“Support” is not the test; it’s one piece of the test
The lawsuit leans heavily on “support”—money spent on “food, healthcare, housing, training, grooming, transportation.”
But support is only one element among several. Even the IRS’s simplified rules for dependents are a set of gates: citizenship/residency status, qualifying child vs qualifying relative, and limitations on who can claim whom.
This is not accidental. Support alone is a terrible organizing principle for a tax system, because support is easy to claim and hard to police at scale.
If “I paid more than half of its expenses” becomes the controlling standard, dependency becomes a contest of receipts and storytelling—and the IRS becomes the nation’s largest lifestyle auditor.
The December 2025 court order makes the same point from a judicial posture: if the claims are likely to be dismissed, dragging the parties through broad discovery is wasteful.
Translation: this isn’t a close call.
The floodgates problem: dogs don’t create a principle—just a preference
Here is the article’s core point, and I’ll make it without flinching:
If you redefine dependent to mean “living being that relies on me financially,” you have not created a dog policy. You have created a tax shelter for ownership.
And then you have to answer the only question that matters:
Where does it end?
If dogs qualify, why not cats?
If cats qualify, why not parrots?
If parrots qualify, why not backyard chickens?
And backyard chickens are not hypothetical. Reporting in recent years has documented how large this category has become; one widely cited estimate puts backyard chicken ownership at roughly 11 million U.S. households, with many owners treating them as pets rather than agriculture.
Now add hobby farms—goats, sheep, pigs, miniature horses, and yes, full-size horses. If you’ve never looked at a monthly feed bill for a horse, trust that the word “dependent” starts sounding very attractive the moment you do. And it doesn’t even stop at mammals. If the standard is simply “a living creature I financially support,” then what’s the limiting principle for fish? Goldfish, bettas, tropical tanks—people buy them as pets every day, they rely entirely on human care, and (unlike a horse) they’re cheap enough that you could scale the claim absurdly fast. Under a “support” theory, a person could stock a tank with dozens of fish and argue each one is a dependent because each one has ongoing food and habitat costs, however small.
Under a “support” theory, an enthusiast could acquire animals purely to manufacture credits. Not because people are evil—because incentives work. The IRS and courts understand this, which is why “fraud and abuse” shows up even in the court’s early stay order.
The complaint tries to frame this as unfairness—some households get credits and others don’t.
But that only works if you ignore the underlying design: human dependency provisions are not a generic reimbursement program for caring about something.
They are a social-policy framework tied to humans, households, and civil status.
The identification problem: dependents are tracked; pets are not
Even if Congress wanted to open the door (it doesn’t, but pretend), the current system is built around identification requirements that assume a dependent is a person.
For example, the Child Tax Credit statute contains a “qualifying child identification requirement”:
And IRS guidance for dependents repeatedly treats a Taxpayer Identification Number (SSN/ITIN) as basic infrastructure—something you can’t claim without.
Dogs don’t have SSNs. Chickens don’t have ITINs. Horses don’t have taxpayer identification numbers.
You can build an animal registry, sure. But then you’ve quietly admitted the truth: this would require an entirely new statutory framework, an enforcement bureaucracy, and a fraud model that would make “identity theft” look quaint.
So when the lawsuit says “this is urgent,” it’s accidentally right—for the opposite reason. It is urgent to not let courts convert a tightly defined legal term into a sentimental catch-all.
“But service animals get tax treatment!”—yes, and that proves the opposite of what the lawsuit wants
The complaint argues that because service animals can be treated as deductible medical expenses, dogs aren’t inherently disqualified from tax benefits.
That’s true—and it’s also the strongest argument against a “pet dependent” theory.
Because service-animal-related deductions exist for a specific, narrow reason: they are tethered to medical care for a person. IRS Publication 502 explicitly contemplates costs for a guide dog or other service animal as part of medical expenses, including certain care costs.
That is not “the dog is the dependent.” It is “the person has qualifying medical expenses.”
Similarly, the court’s December 2025 order itself notes there are limited pet-related deductions in other contexts (service animals, business deductions, foster/charitable contexts), even while rejecting the dependent idea.
So yes: the tax code sometimes recognizes animals.
But it recognizes them instrumentally, attached to human categories that already exist in law: disability accommodation, business activity, charity.
That is how policy boundaries look when they’re not being set on fire.
The deeper principle: children are not just “expensive”; they are part of the civic system
Now we get to the part that matters philosophically, and that gets lost in the dog-cuteness haze.
Human-dependent rules are not built on the premise “anything that costs money deserves a credit.” They are built on civil status and social policy.
Even if you strip it down to cold economics: programs and credits tied to children are linked to families, labor markets, and the future tax base. Children grow into wage earners; they become taxpayers; they participate in the civic system in a way animals do not.
That doesn’t mean animals are valueless. It means they are not citizens-in-waiting. They do not (and cannot) become W-2 employees. They don’t file returns. They don’t earn taxable income in the manner the system is organized around.
So when someone tries to hijack the word “dependent” to mean “emotionally important household member,” they are not asking for fairness. They are asking for the tax code to validate a relationship category the law does not recognize in that way.
And even the complaint’s own wording gives the game away: it argues dogs should be treated as “quasi-citizens.”
That is not a tax argument. That is a redefinition of personhood smuggled into a 1040.
If someone wants a “pet credit,” they should argue for a pet credit—openly, legislatively, and with limits
There is a coherent policy debate hiding under this lawsuit:
Should society encourage pet adoption?
Should there be targeted incentives for spay/neuter, vaccinations, or shelter adoptions (which reduce public costs)?
Should disability-related accommodations for service animals be broadened or clarified?
Should working animals in specific roles be treated differently?
Those are real questions. They are also questions for legislatures, not courts.
Because “make my dog a dependent” is not policy. It’s a shortcut. And shortcuts are exactly how you end up with the “32 chickens = 32 dependents” scenario that breaks the system’s credibility.
Once credibility is gone, compliance erodes. And when compliance erodes, everyone pays more—or gets less.
Closing: this dog won’t hunt (and the court already knows it)
The court’s December 2025 order is not a final merits decision, but it signals what anyone reading the statute already knows: the dependent category is human by design, and a sweeping rewrite through litigation is unlikely to survive.
This isn’t about whether a dog is dependent in the common-sense meaning of the word. Of course pets are dependent.
This is about whether a legal term of art should be converted into a universal reimbursement mechanism for privately chosen obligations.
I love my dog. I also pay for livestock. If the federal government wants to subsidize animal ownership, it can propose that openly, with caps, definitions, enforcement rules, and an honest price tag.
But calling a golden retriever a dependent is not a clever hack.
It’s a principle-free rewrite with no stopping point—and it turns the tax code into a punchline.
