dissenting:

Here is what I thought the majority was saying when I wrote my original dissent in this case: Because Kahn and Caldwell would be deemed common law spouses in some states, they had to be treated by the INS as married even in those states that do not recognize common law marriage. 1

Evidently I was mistaken. In what purports to be a clarification of its opinion, the majority now severs even that feeble tether between law and family. Untroubled by the novelty of calling two people not related by either blood or law “family,” the majority rejects not only state law definitions of marriage, but the notion of marriage altogether. Borrowing from the law of nowhere, the majority commissions the INS to invent a federal definition of “family,” one that turns not on any biological or legal connection, but on whether people have a “warm and continuing relationship” that is “close and deeply emotional.” Maj. op. at 1415 n. 5. Doubtless *1421 this will make for more piquant transcripts in immigration cases, but is it law? The majority’s freestyle adoption of a national definition of “family” falls well outside our competence, and is bad policy to boot.

I

Nothing in the Immigration and Nationality Act says the INS must take family ties into account when it determines whether an alien who has committed a crime may be granted a waiver of deportation under section 212(e). Instead, Congress gave the Attorney General and her delegates discretion over deportation, an authority courts have interpreted broadly, Ayala-Chavez v. INS, 944 F.2d 638, 641 (9th Cir.1991), and the exercise of which is subject to very limited judicial review. See Abdelhamid v. Ilchert, 774 F.2d 1447, 1460 (9th Cir.1985). In exercising this broad discretion, the INS has given life to the text of the statute by balancing a petitioner’s positive equities against negative factors, a methodology we have upheld as rational. Id. at 641 (applying Chevron v. Natural Resources Defense Council, 467 U.S. 837 , 104 S.Ct. 2778 , 81 L.Ed.2d 694 (1984)). In cases such as Matter of Marin, 16 I & N Dec. 581 (1978), the INS has elaborated factors relevant to the weighing of the equities, among them a petitioner’s “family ties” to the United States. Id. at 584 . 2

Given that “family ties” counts as a positive equity only because the BIA itself says so, our review of how the Board interprets this term is extremely limited. It is not, as the majority suggests, merely the deference due under Chevron , which applies to the agency’s interpretation of statutory language and already gives the agency much leeway. An agency’s interpretation of a standard it has itself adopted is entitled to even greater deference, much like that given to an agency’s interpretation of its own regulations:

Udall v. Tallman, 380 U.S. 1, 16 , 85 S.Ct. 792, 801 , 13 L.Ed.2d 616 (1965) (emphasis added). The agency’s interpretation of its own standard may only be rejected if it is “arbitrary and capricious,” Montana Power Co. v. EPA, 608 F.2d 334 , 344 (9th Cir.1979) — i.e., totally devoid of rationality. 3

So the question we must ask ourselves is this: Had the members of the BIA taken leave of their senses when they decided to treat applicants for section 212(c) relief as having “family ties” in the United States only if they could show familial relationships recognized by the applicable law of domestic relations? 4 I would have thought this question could be answered only one way. My *1422 colleagues, however, reject the INS’s approach as totally irrational because they believe that what really matters in a relationship is warmth of emotion and depth of commitment, not outmoded concepts such as marriage and legal obligation. This is gross overreaching. One could, I suppose, refer to an unmarried couple as “family” because they share living.expenses and a bed. 5 But can we possibly say this is the only rational view of what constitutes a family? There is surely still a legitimate view in our society— even way out here in California — that two otherwise unrelated people become a family only when they undertake the legal obligations that go along with being married.

At the heart of the majority’s opinion lies a significant policy disagreement with the INS about what types of relationship merit deference in considering an application for section 212(c) relief. The INS has — quite reasonably in my view — decided to limit consideration to the relatively narrow class of status relationships that are already invested with myriad legal and moral consequences. See Kozinski original dissent at 1417-18. The majority believes that the class of relationships deserving favorable consideration by the INS should be greatly expanded to encompass all connections between people who five together and have deep feelings for each other.

Whence comes our authority to cram this policy down the BIA’s gullet? The statute certainly gives us no support; I challenge the majority to point to anything in the uncommunicative text of section 212(c) that precludes the BIA from recognizing positive equities only where the relationship is between people related by blood or marriage. 6 Congress entrusted the administration of the statute to the INS; the policy choices that govern are the agency’s, not ours. See Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 416 , 91 S.Ct. 814, 823 , 28 L.Ed.2d 136 (1971); California v. FCC, 905 F.2d 1217 (9th Cir.1990).

II

As I noted in my original dissent, ease of administration is an important consideration in judging the rationality of an agency’s implementation of the statutory program entrusted to it. See EEOC v. Commercial Office Prods. Co., 486 U.S. 107, 124 , 108 S.Ct. 1666, 1676 , 100 L.Ed.2d 96 (1988) (affirming EEOC’s rule in part because it is “easily administered”); Stowell v. Secretary of HHS, 3 F.3d 539, 544 (1st Cir.1993) (a “cardinal reason why deference is due” is because the agency’s interpretation “facilitates ... administration”); cf. Elden v. Sheldon, 46 Cal.3d 267 , 250 Cal.Rptr. 254, 259 , 758 P.2d 582, 587 (1988) (refusing to extend cause of action for negligent infliction of emotional distress to cohabiting fiancees, because only legal marriage offers a “sufficiently definite and predictable test to allow for consistent application from case to case”). The majority’s original approach already posed monumental problems of administration for the INS. See Kozinski original dissent at 1418- *1423 19. The “new and improved” standard makes matters -worse.

How, pray, will Immigration Judges peer into the hearts and minds of couples to see if they have the requisite “close and deeply emotional” relationship? How does the BIA — or do we, for that matter — review a determination that there are (or are not) sufficiently warm feelings among the parties to make them “family” under the majority’s “flexible, uniform standard”? Maj. op. at 1415-16. Depth of feeling is easily faked; it’s not all that difficult to bring in chums willing to swear that you’re every bit as close as a married couple. Deciding whether two people are married under the law of their state is generally a eut-and-dry matter, and much harder to he about.

The adoption of a “flexible, uniform standard” may also have other unpleasant consequences. If we are abandoning state law in favor of a uniform national standard, all romantic relationships will have to be judged by this new standard, even if the parties happen to be married under state law. 7 What about those married couples whose relationship has cooled off a bit, maybe grown stale with age? I am more than slightly uneasy about the intrusion into marital privacy caused by giving the INS a charter to investigate whether married couples measure up to some “uniform, flexible standard” of intimacy. 8 The majority has taken a simple determination — whether two people have a certain legally-recognized relationship — and turned it into an encounter group session. I should think the INS is entitled to decline this task on grounds of decorum alone.

Conclusion

Whatever one might say about the relative merits of the BIA’s and the majority’s approaches, it seems to me the BIA is perfectly within its rights in declining to adopt the majority’s touchy-feely theory of human relations. Some may agree with the Beatles that “All You Need is Love,” but in our society people still regularly get married as a means of formalizing their commitment to each other. The majority shrugs off marriage as an idle gesture signifying nothing, contrary to the great significance it continues to have both in our hearts and in our laws. 9 I cannot agree with the majority that the BIA was irrational in holding that marriage — real, state-sanctioned, old-fashioned marriage— matters. By trying to patch up a shaky opinion, the majority has only made things worse — much worse. I continue to dissent, only more so.

. Because the majority continues to refer to common law marriages and quotes Caldwell describing their relationship as “like [a] common-law” marriage, maj. op. at 1413-14 it's unclear whether it has abandoned its original approach. I therefore leave my earlier dissent on the books, just in case.

. "Inherent in this discretion [to grant or deny waiver of deportations] is the authority of the Attorney General and his subordinates to establish general standards that govern the exercise of such discretion, as long as these standards are rationally related to the statutory scheme.” Ayala-Chavez, 944 F.2d at 641 .

. The BIA may not simply disregard its own standards and policies in a particular case without giving a reason for doing so. See Israel v. INS, 785 F.2d 738, 740 (9th Cir.1986). But this is not a case where the BIA disregarded its own precedents and struck out in a new direction; everyone agrees that "family ties" is the applicable standard. The BIA might fall foul of Israel if it refuses to grant outstanding equities to a married couple. But see p. 1423 & n. 8, infra. But surely Israel is not violated by saying that “family” in "family ties” means what most people think it does.

.The majority head-butts a straw man when it points out that each of the statutes described in my original dissent “either expressly states or strongly implies that its application is controlled by a person's status under state law." Maj. op. at 1414-15. I certainly don't argue that section 212(c) requires the INS to consider state law when it determines family status. My point is that it’s more than rational for the INS to do so in light of a large body of federal law where Congress has "either expressly state[d] or strongly implie[d]” that state domestic relations law controls. Id. at 1414-15. I find it positively bizarre to say the INS acts irrationally when it emulates Congress's policy in directly analogous contexts.

. Although I think it's misguided, I don't go so far as to say that the majority's standard is irrational. Presumably rational judges elsewhere have arrived at a similar conclusion: In Dunphy v. Gregor, 136 N.J. 99 , 642 A.2d 372 (1994), the New Jersey Supreme Court allowed a fiancee, whose relationship with the decedent was sufficiently intimate, to seek damages for negligent infliction of emotional distress. But as the vigorous (and in my view totally correct) dissent in Dunphy pointed out, the fountainhead of bystander liability, the California Supreme Court, has refused to extend the cause of action for negligent infliction of emotional distress to unmarried cohabitants. See Elden v. Sheldon, 46 Cal.3d 267 , 250 Cal.Rptr. 254 , 758 P.2d 582 (1988). The dissenter in Dunphy and the majority in Elden — respected (and sane) jurists all — are proof enough that the INS’s view is not the product of dementia.

. The majority props up its critique of the BIA's interpretation with a suggestion that the INS is improperly subjugating federal law to state law. Piffle: How can it possibly offend principles of federal supremacy for a federal agency to choose to rely on the existence or absence of legally enforceable relationships? This does not, as the majority claims, give-state law "a conclusive veto power” over the BIA’s decision; it merely recognizes that the existence of legally enforceable familial obligations is a reliable measure of a petitioner's ties to the United States.

. The New Jersey Appellate Division certainly saw the force of this logic:

Dunphy v. Gregor, 261 N.J.Super. 110 , 617 A.2d 1248, 1254 (1992) (emphasis added).

. The California Supreme Court in Elden expressed similar concerns:

758 P.2d at 587 .

."Drawing the line at marriage conforms to the general expectation of society ... that spouses should be treated differently than unmarried cohabitants,” Dunphy, 642 A.2d at 382 (Garibaldi, J., dissenting); failing to so draw the line "will cause confusion in many other areas of law in which spouses continue to be treated differently.” Id., Id., 642 A.2d at 381 (Garibaldi, J., dissenting). Amen.