On May 8, 2026, the Board of Immigration Appeals issued another precedential decision attacking gender-based asylum claims: Matter of V-A-B-. I just finished reading it, but here is what I take from it – and where I think the litigation is now headed.

By itself, the decision is important. But it’s also a bit of a one-two punch after the BIA’s decision last year in Matter of K-E-S-G-. The upshot is that under Matter of K-E-S-G-, gender-based social groups are never a thing (according to the Board). And under Matter of V-A-B-, groups tied to domestic violence and relationships one can’t leave are also, now, not a thing. At least not at the Board. Simply put, the Board is clearly building a framework where no gender-based particular social groups can ever exist. The fight remains in the Circuit Courts. And I think there is still quite a bit to fight over.

A lot of people remember Matter of A-B- I as the decision that destroyed domestic violence asylum claims. It was Jeff Sessions’ magnum opus in attacking domestic violence victims. It was 31 pages, sweeping in its criticism. But if you go back and actually read the case carefully, the decision did basically nothing doctrinally. There were pages and pages of dicta about criticizing private violence claims and warning about expanding asylum law. But the actual holding was case-specific; it didn’t change the law (the DOJ admitted as much in Grace v. Barr). The real work that Matter of A-B- I did was communicate to Immigration Judges that they weren’t supposed to be granting asylum anymore, even if the law hadn’t changed at all.

They tried again in Matter of A-B- II, right at the end of Trump I. But a few months later, Attorney General Garland outlined how Matter of A-B- had done little more than “create confusion and discourage careful case-by-case adjudication.” Trump later revived Matter of A-B- last year. What has followed is a withering attack, at least at the agency level, on asylum claims by domestic violence victims.

In essence, K-E-S-G- and V-A-B- put a shiny bow on Trump’s ultimate project, started back in 2018.

K-E-S-G- attacks gender-alone social groups and gender-plus-nationality social groups. According to the Board, groups like “Salvadoran women” or “Mexican women” are too broad, too internally diverse, too diffuse, and too disconnected to qualify as a “particular social group.”

V-A-B- attacks the next category of claims, the ones that had been recognized by Matter of A-R-C-G–. The proposed group was “married Mexican women who are unable to leave their relationship.” The Board rejected it as circular and lacking particularity, claiming that “unable to leave” was a description of the abuse itself, and the particular social group can’t be defined by the persecution one is fleeing. Historically, though, “unable to leave” did not just refer to physical violence or fear. It also encompassed economic dependence, children, religious pressure, social stigma, and broader patriarchal structures that trapped women in abusive relationships. V-A-B- collapses all of those concepts into persecution itself.

That is, if K-E-S-G- says gender-based groups are out, V-A-B- says the fallback formulations are out too.

That is a much broader project than the original A-B- decisions ever really accomplished. And after Loper Bright, I am not at all convinced the federal courts are going to accept the Board’s framework anymore.

The Board’s “Social Distinction” House Of Cards Has A Chevron Problem

The most important issue in these cases may not actually be asylum law at all. It will almost certainly be administrative law and what to do now that the Chevron doctrine is gone. Let me explain.

The statute protects people persecuted on account of “membership in a particular social group.” That phrase is never defined in the statute or in any agency regulation (In Matter of A-B- III, Garland’s DOJ promised to issue regulations defining the term, but it never happened).

In its landmark decision in Matter of Acosta, the BIA explained that particular social groups get their status as such from their “immutability.” To really be a social group, the group’s members had to have a trait or traits that they couldn’t or shouldn’t have to change. But over time, the Board created a framework requiring more:

• immutability;
• particularity;
• and social distinction.

The modern formulation mostly comes from Matter of M-E-V-G– and Matter of W-G-R-.

But those requirements are not in the INA. They are creations of Board precedent. This framework was built during the high-water mark of Chevron deference, when courts often deferred heavily to agency interpretations of ambiguous statutes. A couple circuits had rejected these requirements outright. One of them was the Seventh Circuit in Gatimi v. Holder, which, coincidentally, was my first Circuit Court appeal.

The Board responded in Matter of M-E-V-G-. That decision explicitly leaned on National Cable & Telecommunications Association v. Brand X Internet Services and the idea that the Board could impose a nationally controlling interpretation of “particular social group” even where some circuits had previously gone in different directions. So the Board said it could require “social distinction” (which it previously called “social visibility”) and the circuits would have to fall in line. And, for the most part, the circuits deferred to that view under Chevron.

Before Matter of M-E-V-G-, it wasn’t all that certain that “social distinction” or “visibility” could even be required. A lot of circuit courts hated the social visibility concept. Judge Posner’s opinion in Gatimi v. Holder was withering. Other circuits were skeptical too. Because the doctrine always had a slippery quality to it.

Distinct to whom?
The persecutor?
The government?
Society at large?
An expert witness?

How do we measure “social distinction” in the first place? The Board’s answer has never been especially coherent.

Today’s decision in V-A-B- sort puts the final touches on the Board’s project. It is saying that immutability alone is insufficient. And that’s not hyperbole; V-A-B- says that directly. The Board insists that all three requirements must be satisfied: immutability, particularity, and social distinction. And it now insists that gender-based groups and domestic violence-based groups aren’t going to meet that standard.

That matters as we take the fight to the Circuit Courts, because those extra requirements are not statutory text. They are judge-made agency doctrines layered on top of a statute that simply says “particular social group.”

Before Chevron died, courts often treated these questions as largely settled. The Board got deference. That world no longer exists.

Loper Bright Changes The Conversation

Then came Loper Bright Enterprises v. Raimondo. Circuit courts are no longer required to reflexively defer to the Board’s interpretation of “particular social group.” Courts are now free to decide for themselves whether “social distinction” and “particularity” are actually faithful interpretations of the statute or whether they became agency-created limiting doctrines designed to narrow asylum eligibility.

And I think there is a serious argument that the Board’s framework is vulnerable.

The Acosta immutability framework has a much stronger connection to the text of the INA and the Refugee Convention than the later M-E-V-G– framework does. That decision spells out how the immutability test was developed historically in asylum and refugee law. Acosta tried to identify what makes a social group a social group. The later cases increasingly read like efforts to impose limiting principles on a category the Board worried was becoming too broad (oh no, the floodgates!).

That is especially true with “social distinction.”

And in this moment, when over 100 Immigration Judges have been fired by Trump (most of the judges with reasonable asylum approval rates) and judges have accused the agency of ordering them to rule a certain way, it is hard not to wonder whether “social distinction” has simply become a convenient doctrinal tool for reaching predetermined outcomes while making sure the decision survives appellate review should the noncitizen appeal.

And once Chevron disappears, courts are free to ask a question they’ve been avoiding since M-E-V-G- was published in 2018: why should they continue accepting doctrinal requirements that do not actually appear in the statute?

The Board’s Reasoning About Sex Is Surprisingly Weak

K-E-S-G– contains one passage that really stood out to me.

The Board says that recognizing women as a particular social group would effectively create “sex” as a separate protected ground even though Congress did not explicitly list sex in the refugee statute. I don’t think that reasoning holds up very well.

The whole purpose of “particular social group” was to function as a catch-all category. The Board itself has recognized social groups built around characteristics not specifically listed in the statute:

• family relationships;
• tribes;
• clans;
• sexual orientation;
• former occupations;
• disability;
• and many others.

So the fact that Congress did not separately list sex does not really tell us much. After all, Congress also did not separately list “family membership,” yet family-based social groups are among the most widely recognized PSGs in asylum law (and always have been).

And there is another problem with the Board’s reasoning. The Board repeatedly describes women as too broad, too internally diverse, and too diffuse to qualify as a social group. But asylum law does not require internal homogeneity for the other protected grounds. The Board itself acknowledged back in 2006 in Matter of C-A– that social groups do not require internal homogeneity, and none of these newer decisions purports to overrule that principle.

Nor would such a rule make any sense. Religions are internally diverse. Nationalities are internally diverse. Political movements are internally diverse. The statute does not require every member of a protected category to look alike, think alike, or live alike.

The Board seems to be applying a much harsher coherence requirement to “particular social group” claims than it applies to the statute’s other protected grounds. And it is difficult to ignore where that heightened scrutiny tends to appear: cases involving women, domestic violence, and claims the agency has spent the better part of a decade trying to narrow.

The Circularity Analysis In Matter of V-A-B– Goes Further Than Earlier Cases

The Board also expands the anti-circularity doctrine in a problematic way. Earlier circularity cases usually involved groups explicitly defined by persecution itself, like “people targeted by gangs” or “victims of extortion.” The idea behind the circularity problem is that persecution must be “on account of” something specific about the person, like membership in a group. But the group needs to be identifiable on its own, not just by the fact that the group gets persecuted. If you didn’t have that backstop, you could have a particular social group called “people who are going to get persecution” and the requirement of nexus imposed by the statute would mean nothing at all.

But historically, “unable to leave a relationship” meant a lot of things:

• economic coercion;
• religious pressure;
• family pressure;
• social stigma;
• children;
• isolation;
• patriarchal structures;
• legal obstacles;
• threats of violence.

Those are immutable traits, but they’re not the “persecution” itself. V-A-B- pretends that none of these exist. It largely collapses all of that into persecution-created inability.

The Board treats “unable to leave” almost as synonymous with “afraid to leave because of abuse.” That narrows the concept substantially compared to how many courts, advocates, and even prior agency decisions had historically understood these claims. And the BIA was clear in Matter of W-Y-C- that it has no authority to reframe the proposed group into some other group on appeal.

This distinction is super important, because many abusive relationships are not physically inescapable in the literal sense. No prior BIA or circuit court decision had interpreted ‘unable to leave’ to mean literally, physically unable to leave. The barriers are often social, economic, cultural, religious, and psychological. Sometimes the person technically could leave. The real question is whether leaving is realistically possible without catastrophic consequences.

That is what made Matter of A-R-C-G– significant in the first place. It recognized that coercive control can trap people in relationships in ways that are far more complicated than simply being afraid of future violence.

V-A-B- strips that reality out of the analysis. The decision reframes “unable to leave” as little more than a description of the persecution itself. Once the Board does that, the group becomes “circular” and therefore invalid.

That is a major doctrinal narrowing move, even if the Board presents it as nothing more than straightforward application of existing doctrine. It’s not.

The Board Keeps Talking About Size Without Calling It Size

We also have a size problem. K-E-S-G– repeatedly insists that numerical size alone is not disqualifying. But the opinion spends pages describing these groups as “broad,” “diffuse,” “disconnected,” and encompassing large swaths of society.

At some point the distinction starts collapsing in on itself.

Officially, the doctrine says large groups can still qualify. The Eighth Circuit has recognized All Somali Women as a particular social group, for example. But in practice, once a proposed group encompasses a substantial segment of society, the Board often starts treating size itself as evidence that the group lacks “particularity.”

Today’s Decision Quietly Signals Skepticism Despite Factfinding Limitations

Another thing stood out to me in today’s decision in V-A-B-. The Board repeatedly criticizes the applicant for lacking corroboration even though the case is not framed as a corroboration or credibility decision. The opinion repeatedly says things like:

• speculative;
• unsupported;
• no credible evidence;
• failed to demonstrate;
• failed to establish.

The Board seems increasingly skeptical of testimony standing alone, especially in domestic violence cases. This is bad. The BIA is not supposed to conduct original factfinding or independently reassess testimony the way a trial court would. But look at how easily the opinion slips into doing exactly that.

The applicant testified about corruption and criminal connections. The Board wanted more.

The applicant testified she could not safely relocate. The Board wanted more.

The applicant testified police would not protect her. The Board wanted more.

That may not sound remarkable at first glance, but many of these cases involve coercive control, private abuse, family violence, and threats that rarely produce clean documentary evidence. If the default posture at the BIA becomes skepticism, the practical effect is a quiet ratcheting-up of evidentiary expectations and a kind of institutionalized disbelief that is not how appellate review is supposed to work.

It is also hard to separate this growing skepticism from the broader transformation of the Board itself over the last several years. The author of V-A-B-, Board Member Keith Hunsucker, was previously an Immigration Judge with one of the lowest asylum grant rates in the country. He did not grant a single asylum case in 2019 while serving as an Immigration Judge before joining the BIA. That background does not invalidate the decision by itself, of course. But it does help explain, to me at least, the increasingly skeptical lens through which many asylum claims, especially domestic violence claims, now appear to be viewed at the agency level.

The Board increasingly seems willing to substitute its own interpretation of the facts while still using the language of deference or “clear error.”

Final Thoughts

There is no question where the Board is trying to take asylum law.

K-E-S-G– and V-A-B- are not isolated decisions. Read together, they are a coordinated effort to narrow “particular social group” doctrine to the point where most gender-based claims become impossible to win at the agency level.

The Board insists this is all just straightforward application of existing doctrine. I do not think that is true. These decisions go substantially further than Matter of A-B- I ever did. Back then, DOJ repeatedly assured federal courts that A-B- did not categorically foreclose domestic violence claims and still required individualized adjudication. The Board is now moving toward exactly the kind of categorical narrowing the government previously denied was happening.

And unlike the original A-B- decisions, K-E-S-G- and V-A-B- are doing real doctrinal work.

But there is another side to this story.

These decisions are arriving at the exact moment when Chevron deference has disappeared. The Board’s modern PSG framework was built during an era when courts often deferred heavily to agency interpretations of ambiguous statutes. That world is gone now.

The phrase “particular social group” still appears in the statute. Congress still never defined it. And the Board’s requirements of social distinction and particularity still do not appear anywhere in the INA itself.

That means the circuit courts are now free to ask a question they often avoided during the Chevron era: whether the Board’s framework is actually the best reading of the statute in the first place.

I suspect that question is coming sooner rather than later.

And when it does, the next chapter of this fight probably will not be written by the Board of Immigration Appeals.