concurring in part and dissenting in part.
Had the district court’s opinion been allowed to stand as originally filed, JSN and MKN—the two younger Neumann children who were ordered to be returned to Mexico—would clearly have faced an “intolerable situation.” For one thing, the district court failed to enter any specific orders concerning their return, directing only that they were to go to Mexico on a certain date, but not back to their father, and that their mother was to arrange for their transportation to Mexico, but not necessarily accompany the children or supervise them there.1 Because neither side of the family had relatives in or near Mexi-*485eo City, and no other' arrangements for their protection or supervision were made, one can only wonder about their care and safety upon arrival. The district judge ultimately determined that the 13- and 14-year-old brothers were not sufficiently mature to have their preference against returning to Mexico honored, but the court apparently considered them mature enough to manage on their own once they got there.2
The omissions in the district court’s order are startling, to say the least. According to the Federal Judicial Center’s publication on handling Hague Convention abduction cases, “[w]hen an order for the return of a child is made, courts should focus on enforcement, specificity, and the safety of the child” and “clearly state the mandated time, place, and details of the child’s return.”3 District courts are advised to invoke “mirror-image orders,” which “are entered in both the courts of the state hearing the petition and the courts of the child’s habitual residence,” or to “direct counsel for the parent requesting return to obtain a safe harbor order from the courts of the habitual residence.” The latter are “designed to avoid severe and immediate physical or psychological harm to the child as a result of the conditions of return” and may be required as a prerequisite of an order of return.4 None of this was accomplished by the district court order in this case.
Fortunately for all concerned, we have now learned, through statements elicited at oral argument, that Steven Neumann is no longer in Mexico because his assignment from his employer ended some time last year. According to his attorney, he is now back in Michigan, apparently in Wayne County, as are Julie Neumann and all three children. Despite this fact, the majority presumes that even though “neither parent currently resides in Mexico,” the children may be required to end up there, if the district court again rejects Julie Neumann’s “grave risk” defense. I gather that this conclusion is based on an assumption that Mexico, as the country of their “habitual residence” in 2014, is the only jurisdiction in which the children’s custody can be litigated under the Hague Convention. But that assumption is flat wrong. As the Ninth Circuit has noted:
The Convention’s principal remedy is the return of the abducted child. However, the Convention does not make clear to what country a child must be returned. The Preamble recites the Convention’s goal as the return of children “to the State of their habitual residence.” However, the actual text of the Convention is silent as to where the child should be returned. Article 12 merely provides that a wrongfully removed child should be “returned ... forthwith.”
The Convention’s official commentary reveals that this silence was intentional. ■See Elisa Perez Vera, Explanatory Re*486port ¶ 110, in 3 Hague Conference on Private International Law, Acts and Documents of the Fourteenth Session, Child Abduction 459-60 (1982). The commentary explains that the Convention rejected a proposal that would have required a child to be returned to his habitual residence. Id. The Convention was concerned that such a proposal would prove “inflexible” when the petitioner moves from the State of the child’s habitual residence post-abduction. Id. In other words, the Convention did not provide that a child be returned to his pre-abduction habitual residence if the petitioner had relocated to a different country. The Commentary states that the Convention intended that the child be transferred to the petitioner’s new residence in these circumstances. Id.
The State Department’s commentary on the Convention contains a similar discussion. “The Convention does not technically require that the child be returned to his or her State of habitual residence, although in the classic abduction case, this will occur. If the petitioner has moved ... the child will be returned to the petitioner, not the State of habitual residence.” Hague International Child Abduction Convention; Text and Legal Analysis, 51 Fed. Reg. 10404 (Mar. 26, 1986).
Von Kennel Gaudin v. Remis, 282 F.3d 1178, 1182-83 (9th Cir. 2002) (citations omitted) (alteration in original) (emphasis added).
The admonition to return the children to their father and “not the State of habitual residence” creates an apparent legal impasse in this instance, given that the district court specifically denied return to the father. However, under the facts, of this case, there is a solution to this dilemma, because the Von Kennel Gaudin opinion also holds that “[t]he Convention cannot be invoked when the petitioner moves permanently to the same country in which the abductor and the children are located.” Id. at 1183 (emphasis added); accord Witherspoon v. Orange County Dept. of Social Servs., 646 F.Supp.2d 1176, 1182 n.2 (C.D. Cal. 2009), March v. Levine, 2006 WL 3805665, at *4 (M.D. Tenn. 2006). In other words, the case becomes moot if both parents are in the United States. See Von Kennel Gaudin, 282 F.3d at 1183 (citing IRS v. Pattullo, 271 F.3d 898, 901 (9th Cir. 2001), for the proposition that (“[i]f an event occurs while a case is pending on appeal that makes it impossible for the court to grant any effectual relief whatever to a prevailing party, the appeal is moot and must be dismissed”)); see also Ford v. Wilder, 469 F.3d 500, 504 (6th Cir. 2006) (“Simply stated, a case is moot when the issues presented are no longer ‘live’ or the parties lack a legally cognizable interest in the outcome.”) (quoting Powell v. McCormack, 395 U.S. 486, 496, 89 S.Ct. 1944, 23 L.Ed.2d 491 (1969)). Indeed, “[Regardless of whether the parties raised the issue of mootness, ‘our first inquiry on appeal must be whether this case is moot.’ ” Id. (quoting McPherson v. Mich. High Sch. Athletic Ass’n, Inc., 119 F.3d 453, 458 (6th Cir. 1997) (en banc)).
Dismissal of this appeal as moot is not only the legally correct result in this case; it would also be the most efficient way to get the parties into the court in which their long-running custody dispute can finally be heard and resolved appropriately. Certainly, nothing can be gained, as the majority would have it, by a remand and an order to the district court to rehear the same evidence that ultimately must be presented in the Wayne County Circuit *487Court’s Family Division.5 If, however, the majority nevertheless insists on a remand, I would suggest that the district court first re-examine an issue that will not be replicated in state court, ie., the question of JSN’s and MKN’s objections to being returned to Mexico and/or to their father against their will.
The district judge interviewed the two younger children in chambers with only a law clerk present and determined that they had “a preference” to remain in Michigan but no actual “objection” to returning to Mexico, or any “fear of harm” from their father. Of course, we cannot know how subdued and polite the two children may have been in the presence of the judge and the absence of any supportive family members, but the record is replete with their statements to Dr. Haynes, to two different therapists, and to their mother’s pastor regarding their fears about returning to Mexico and their apprehension that they would not be “safe” with their father—all of which the district judge discounted because the' statements were made to “third-parties” and not “directly” to the judge in chambers. Moreover, the district court’s determination that the two children had to be returned to Mexico because they expressed only a “preference” to live in Michigan and not an “objection” to returning to Mexico is not only belied by the record but is also an example of deciding a question of law on the basis of purely superficial semantics. In point of fact, the Text and Legal Analysis of the Hague Convention developed by the Department of State uses the terms interchangeably.6
The district court’s alternative finding that the two children were not mature enough to make a decision about their residence, based on Dr. Haynes’s report, also does not hold up. In a four-paragraph summary of his views on “the extent to which the minor children [c]an make mature decisions about returning to Mexieo,” Haynes indicates that “all three children continue to be strong and firm about not wanting to return to Mexico at this time” and that they appear to have made independent judgments and “come to their own views regarding ... return to Mexico.” He notes that “[f]rom a psychological standpoint, it would be inappropriate, anxiety-producing, and unwise for the children to be split up regarding the issue of Mexico [because] all three would view going to Mexico as a highly coercive action against their will and against their intent, regarding which they have significant fears” that are being “repetitively addressed in psychotherapy.”
The sentence from this part of the Haynes supplemental report quoted by the district court—that “because of the boys’ age, ‘(b)y definition their decisions and cognitive processes on such matters are not mature’ ”—may be no more than a recognition by an expert that an individual’s “cognitive processes” are not fully ma*488ture until age 25. In his initial evaluation, Dr. Haynes found that the three children should not be considered “equally mature” in making decisions about returning to Mexico, because JMN, at 16 years old, “is significantly more mature than the two boys.” But the report does not indicate that the younger two children’s decisions aré intrinsically immature or should be disregarded.
In any event, JSN and MKN are now 14 and 15 years old, beyond the age when most children are permitted to express a preference in ordinary custody cases, and significantly older than the two Simcox children, ages 10 and 12, whose objections to being returned were honored by the district court in Simcox v. Simcox, 499 F.Supp.2d 946, 952 (N.D. Ohio 2007) (Simcox I), rev’d on other grounds, Simcox v. Simcox, 511 F.3d 594 (6th Cir. 2007) (Simcox II). And, although the district judge in that case declined to consider the views of an eight-year-old Simcox child, the court nevertheless acknowledged that children as young as eight years old have been found of sufficient age and maturity to object to repatriation. Id.; see also, e.g., Anderson v. Acree, 250 F.Supp.2d 876, 883-84 (S.D. Ohio 2002) (eight-year-old found to be of sufficient age and maturity); Blondin v. Dubois, 238 F.3d 153, 166 (2d Cir. 2001) (same).
Finally, I cannot agree with the majority’s decision to provide the district court with another opportunity to consider the question of “grave risk of harm” to the children if they are sent back to Mexico. It seems clear that Julie Neumann has already discharged her burden of proof in this regard. To begin with, the district court’s decision to send JSN and MKN back to Mexico but not back to their father can be seen only as an implicit determination that the children’s return even to Steve Neumann’s temporary custody posed a grave risk of physical or psychological harm to them. In addition, the district court took what the record establishes as a compelling history of alcohol-fueled domestic abuse of their mother and portrayed it as nothing more than a he-said/she-said situation, using terminology such as “regardless of ... which version of the December 26 assaults is accepted” and “[e]ven assuming the children did physically intervene when Steven pinned Julie.” Neumann v. Neumann, 187 F.Supp.2d 848, 862 (E.D. Mich. 2016). In almost the same breath, the district court expressed its “conclusion that Steven was not a credible witness.” Id. At that point, the mother’s testimony about what occurred on December 26, especially given its corroboration by the children, should have been accorded the full credit that it deserved.
That testimony painted a stark portrait of domestic abuse, not the least part of which concerned Steven Neumann’s severe but untreated alcoholism, which fueled not only his physical abuse of his wife on December 26, but played a significant role in at least two other physical assaults on Julie Neumann. At Thanksgiving in 2003, Steven Neumann “got real, real drunk” and started hitting her in the head as they drove home from dinner, necessitating a return to Steven Neumann’s parents’ house for help—Julie Neumann and the three children spent the night on the floor in the - family room. In 2005, according to Julie Neumann’s statement to Dr. Haynes, Steven Neumann “grabbed her and threw her against the wall of the kitchen and she hit her head, and did not know if she was knocked unconscious but [ ] she then realized she was on the kitchen floor.” As for the events of December 26, the district court noted that Steven Neumann had been drinking on the two previous days when, in fact, the record shows that he had been drinking heavily since December 21, *489and before that on December 19 and 20, based on a record that he himself supplied as an exhibit. According to Dr. Haynes, Steven Neumann was not a social drinker or someone who simply enjoyed a glass of wine with dinner. Instead, he was a self-described alcoholic and “binge drinker,” who would start drinking early in the day on weekends and would not, or could not, stop drinking once he began. He maintained that his drinking did not affect his work but admitted that it affected his relationship with his wife and children—when drinking, he tended to lose his temper around the family, yelling and cursing at them. During Christmas week in 2014, Steven “was consuming a half gallon of gin, vodka, and tequila combined per day,” and he “additionally consumed beer and wine,” used Valium daily and “also sometimes Ativan.” After Julie and the children fled from the house on December 26 and from the country on December 28, Steven Neu-mann’s two brothers flew down to Mexico City on December 30 to check on him. At that point Steven had been drinking around the clock for four days in an effort, he later said, to “drink himself to death.” His brothers flew Steven home to Michigan to get him admitted for treatment; he arrived at the rehab facility in an intoxicated condition and checked himself out two days later, “against medical advice.”
Although the district court at one point referred to Steven Neumann’s problem as “alleged untreated alcoholism,” id. at 860, there is no question concerning what led up to his physical and verbal assaults on his wife on December 26. Moreover, the problem remained untreated at the time of the hearing, although Steven claimed to be under the care of a psychiatrist in Mexico and to have joined AA there. As it turns out, the physician in question was treating him for depression, not alcoholism, and Steven had been to only three AA meetings in three months. He also claimed that his mother in Michigan was acting as his “mentor” or AA sponsor—an impossibility under AA practice. Although he also claimed to be abstinent at the time of the hearing in the district court, Dr. Haynes predicted that without rehabilitation, Steven Neumann was headed for relapses that would get progressively worse, and that his “substance abuse and its implications” were the “central issue” in the family’s dysfunctional situation.
In determining that the children did not face a grave risk of harm if returned to Mexico, the district court recognized that the issues of domestic abuse and Steven Neumann’s untreated alcoholism were “interconnected,” but then disaggregated them in assessing their impact on the children’s safety. For example, the court pointed to Steven’s excessive consumption of alcohol and testimony from “the children and Julie indicat[ing] that Steven would have difficulty walking and would fall over, his speech would be slurred, and he would otherwise ‘not be himself,’ ” but the court nevertheless found that “Steven’s alcoholism d[id] not appear to pose a physical danger to the children.” That conclusion is clearly refuted by the psychological evaluation that the court ordered, and then largely ignored. It is difficult to imagine how a parent who was frequently “falling-down drunk,” if left in charge of three children, would not present a risk of both psychological and physical harm to those children.
As for the domestic violence committed against Julie, the district court found that the “the incident on December 26 has affected the children” but not to the extent that a “return to Mexico would subject the children to grave risk of purely psychological harm-,” noting that the PTSD the children exhibited immediately following their arrival in Michigan had dissipated by the time they were evaluated eight months *490later by Dr. Haynes. As the district court described the psychological report, it showed only that risk to the children was “significant,” but not that it was “grave.” Actually, the report indicated that in Dr. Haynes’s judgment there were “multiple significant risks” that should be considered cumulatively, including:
• Steven’s refusal to undergo substance abuse treatment, which “raisefd] a clear risk of relapse, with significant results not only to Steve but to the children and to Julie”;
• the extraordinary stresses and risks caused by a contentious divorce;
• the fact that “there has been domestic violence, and it involved injury. Julie has been dealing with this, the children have been dealing with it. It is unclear that Steve has been dealing with it at this point”;
• “The customary practice in ... situations of spousal violence witnessed by the children is that some mental health intervention is necessary ... prior to reunification of the children with the absent parent.”7
The district court’s ultimate conclusion that Julie Neumann had failed to establish a defense was based on its determination that the “alleged abuse” in this case “was only psychological in nature, isolated and sporadic, and it is far from certain that it is likely to reoccur in the future.” Id. at 866. Given this thoroughly superficial analysis, the only obvious certainty here is the district court’s apparent lack of familiarity with the nature and significance of both alcoholism and domestic violence and their harmful effect on children.8 In this case, the violence consisted of Steve Neumann, literally twice the size of his victim, pinning Julie Neumann against a kitchen counter with a knife to her throat, while two of the children pulled at his arm to get him away from her. As the knife clattered to the floor, Steven flung Julie away so violently that she fell against a chair and broke three of her ribs. If this were the only “incident” at issue in this litigation, it would be serious enough to create a defense to “wrongful removal” under the Hague Convention, because it was brutal, it involved two of the children directly and could have caused them injury, and it sent Julie Neumann to the hospital. Although the International Parental Kidnapping Act, 18 U.S.C. § 1204, makes the illegal removal of a child from the United States “with intent to obstruct the lawful exercise of parental rights” a felony, it also creates an affirmative defense if “the defendant was fleeing an incidence or pattern of domestic violence.” 18 U.SiC. § 1204(a),(c)(2) (emphasis added).
There has been a movement to amend the Hague Convention to make fleeing from domestic violence for safety reasons a stand-alone defense to the return of an abducted child. There have also been efforts to revise the convention so that such flight is not a “wrongful removal” in the *491first place. As Judge Richard Posner pointed out in Khan v. Fatima:
The [Hague] Convention was created to discourage abductions by parents who either lost, or would lose, a custody contest. ... The Convention drafters adopted a ‘remedy of return’ ... to discourage abductions, reconnect children with their primary caretakers, and locate each custody contest in the forum where most of the relevant evidence existed. [But] while the remedy of return works well if the abductor is a noncustodial parent, it is inappropriate when the abductor is a primary caretaker who is seeking to protect herself and the children from the other parent’s violence.
680 F.3d 781, 783-84 (7th Cir. 2012) (internal quotation marks omitted) (alterations in original) (quoting Merle H. Weiner, “Navigating the Road Between Uniformity and Progress: The Need for Purposive Analysis of the Hague Convention on the Civil Aspects of International Child Abduction,” 33 Colum. Human Rts. L. Rev. 275, 278-79 (2002) (citations omitted); also quoted in Van De Sande v. Van De Sande, 431 F.3d 567, 568 (7th Cir. 2005)). See also Karen Brown Williams, “Fleeing Domestic Violence: A Proposal to Change the Inadequacies of the Hague Convention on the Civil Aspects of International Child Abduction in Domestic Violence Cases,” 4 John Marshall L.J. 39, 42-45 (2011); Noah L. Browne, Note, “Relevance and Fairness: Protecting the Rights of Domestic-Violence Victims and Lefh-Behind Fathers Under the Hague Convention on International Child Abduction,” 60 Duke L.J. 1193, 1202-05 (2011); Roxanne Hoegger, “What If She Leaves? Domestic Violence Cases Under the Hague Convention and the Insufficiency of the Undertakings Remedy,” 18 Berkeley Women’s L.J. 181, 187-88 (2003); Merle H. Weiner, “International Child Abduction and the Escape from Domestic Violence,” 69 Fordham L. Rev. 593, 634 (2000).
Perhaps the seminal case in this country dealing with domestic violence under the Hague Convention is Walsh v. Walsh, 221 F.3d 204 (1st Cir. 2000), in which the court noted that “both state and federal law have recognized that children are at increased risk of physical and psychological injury themselves when they are in contact with a spousal abuser.” The court also pointed to a congressional resolution, passed in 1990, recognizing that “the effects of physical abuse of a spouse on children include the potential for future harm where contact with the batterer continues” and that “children often become targets of physical abuse themselves or are injured when they attempt to intervene on behalf of a parent.” Id. (alterations omitted) (quoting H.R. Con. Res. 172, 101st Cong., 104 Stat. 5182, 5182 (1990)). The Walsh court concluded that “[t]hese factors are sufficient to make a threshold showing of grave risk of exposure to physical or psychological harm.” Id.
Perhaps the leading case in this circuit on domestic violence under the Hague Convention is Simcox v. Simcox, in which we observed that “the Convention’s purposes would not ... be furthered by forcing the return of children who were the direct or indirect victims of domestic violence.” 511 F.3d at 605 (citation and alterations omitted). The five children in that case “expressed fear of their father” caused primarily by his severe forms of physical “discipline” of his children, but also because he hit and cursed at their mother “on numerous occasions.” Id. at 599. Eventually, Mrs. Simcox grew tired of the abuse, made plans to leave Mexico with her four youngest children, packed up the family car, and drove them across the border to Texas on the way to her family in Ohio. The district court found that the *492children had been wrongfully removed from their habitual residence, but honored the decision of two of the children, ages 10 and 12, to remain with their mother. The court ordered the two younger children, ages four and eight, to be returned to the family residence in Mexico but—concerned for their safety—also ordered that they remain in their mother’s custody and have no contact with their father until “the Mexican Court determines access and visitation rights.” Simcox I, 499 F.Supp.2d at 957. On appeal, we found that Mrs. Simcox could not be forced to return to Mexico and that the risk of harm to the two younger children, if returned to their father, was grave. We then remanded the case to the district court to determine whether there were any conditions that would “be sufficient to ensure the safety of the Simcox children upon their return to Mexico pending the- outcome of custody proceedings.” Simcox II, 511 F.3d at 610. But we also warned that if “the only way in which the children may be protected from harm is for them to remain in the custody of their mother, then it may be necessary to deny the petition.” Id. at 611. On remand, the district court declined to order the children returned to Mexico without their mother and denied the petition. Simcox v. Simcox, No. 1:07CV96, 2008 WL 2924094 (N.D. Ohio July 24, 2008) (Simcox III).
In the Neumanns’ case, the district court should have done the same. Instead, the court minimized the severity of the risk that Julie and the children faced in Mexico, finding that it fell into “the middle-of-the-road” category of the tri-part analysis of Simcox II9 and could not be considered “grave.” Neumann, 187 F.Supp.3d at 866. On appeal, whether there is a grave risk of harm under the Hague Convention is a mixed question of law and fact that is reviewed de novo. In my judgment, when viewed through this lens, the risk of harm to JSN and MKN if returned was at least as grave as that in Simcox, in which the primary risk to the children was the father’s manner of discipline. But in that case, was there longstanding and severe, untreated alcoholism involved? No. Was there abuse of their mother that the children witnessed? Yes, but in Simcox, it did not result in injury requiring medical attention. Nor did it involve the children coming between the parents in an effort to wrest away a knife that their father was holding to their mother’s throat. The gravity of the situation was such that Julie Neumann felt forced to flee the house immediately, to seek medical attention and for her own safety and that of her children. She left Steve Neumann in a drunken stupor, from which he apparently did not awake until the following morning, when he found his family gone. There should have been no question about relief in this case.
As noted above, I would hold that this appeal became moot when Steven Neu-mann returned to Michigan. And although a brief remand may be necessary to confirm his current location, I cannot concur in the remainder of the majority opinion.
. Courts are not authorized by either the Hague Convention or the International Child Abduction Remedies Act (ICARA), 22 U.S.C. §§ 9001-9011, to order a parent to relocate to another country. See Redmond v. Redmond, 724 F.3d 729, 735 n.1 (7th Cir. 2013).
.Despite the district court’s decision to reject the children’s choice to stay with their mother, at another point in the opinion, the court concluded that “the two boys are of an age (13 and 14) where (sic) they are, in many ways, self-sufficient and can manage most of their basic needs without significant supervision,” a finding that certainly evinces maturity. See Neumann v. Neumann, 187 F.Supp.3d 848, 864 (E.D. Mich. 2016).
. Federal Judicial Center, The 1980 Hague Convention on the Civil Aspects of Intemational Child Abduction: A Guide for Judges, 149-50 (2nd ed. 2015), http://www.fjc.gov/public/ pdf.nsf/lookup/Hague-Convention-Guide-Second-Edition-2015 ,pdf/$file/Hague-Convention-Guide-Second-Edition-2015.pdf (hereafter cited as FJC Guide for Judges).
. FJC Guide for Judges 150-51,
. If Steven Neumann’s petition had been dismissed by the district court based on grave risk of harm if the children were returned to their father, the custody and visitation decisions could have been finalized in Michigan over a year ago. Routing the custody determination back through the district court and possibly through the Mexican court system at this late date does nothing more than increase the stress that the Neumanns have been living under for years.
. In a paragraph explaining the limitations on the obligation of return entitled “Child’s preference,” the analysis provides: “The third, unlettered paragraph of Article 13 permits the court to decline to order the child returned if the child objects to being returned and has attained an age and degree of maturity at which it is appropriate to take account of the child’s views." 51 Fed. Reg. 10,494, 10,510 (1986) (emphasis added).
. Dr. Haynes observed in his risk assessment that the "children have expressed feelings of apprehension regarding reunification with their father at this time. They say- they feel unsafe.” Nevertheless, "reunification” with his children was attempted by teleconference. During the meeting, Steven blew up and "fired” the children's therapist, who had been treating them for some time and was in the room with them at the time of the teleconference.
. There are a number of Hague Convention cases in which courts have recognized that even though violence by a parent is not directed toward a child, damage to the child can result. See, e.g., Gomez v. Fuenmayor, 812 F.3d 1005, 1007 (11th Cir. 2016); Khan v. Fatima, 680 F.3d 781, 787 (7th Cir. 2012).
. Simcox II, 511 F.3d at 607-08.