365 Mass. 13 (1974)
310 N.E.2d 112
LUCIO FIORENTINO
vs.
PROBATE COURT & another. CARMEN FERNANDEZ
vs.
GUILLERMO FERNANDEZ.

Supreme Judicial Court of Massachusetts, Suffolk.

September 21, 1973.
March 29, 1974.
Present: TAURO, C.J., REARDON, QUIRICO, BRAUCHER, HENNESSEY, KAPLAN, & WILKINS, JJ.

F. Robert Houlihan (Gershon M. Ratner with him) for Carmen Fernandez.

Sumner D. Goldberg for Lucio Fiorentino.

Raymond H. Young, Special Assistant Attorney General, for the Probate Court & another.

TAURO, C.J.

These cases, argued together, present challenges to the constitutionality of our statutes which, in certain circumstances, require that a person be a resident of Massachusetts for two years before a libel for divorce may be filed. G.L.c. 208, ßß 4-5.[1] This two year durational residence requirement has recently been declared unconstitutional by the United States District Court, District of Massachusetts, Rosado v. Smith (Civil Action No. 72-3361-W),[2] a decision which, while entitled to our respect, is not binding on us. United States ex rel. Lawrence v. Woods, 432 F.2d 1072 (7th Cir.1970), cert. den. sub nom. Lawrence v. Woods, 402 U.S. 983 (1971).

The libellant[3] in each of these cases was married in the State of New York, lived together with his or her spouse in that State, and moved to Massachusetts only after the alleged cause of divorce had occurred in New York. Lucio Fiorentino moved to Massachusetts on April 1, 1972. On December 13, 1972, the Probate Court for Suffolk County granted him a decree that he was living apart from his wife for justifiable cause. On April 10, 1973, Fiorentino executed and attempted to file in the Probate Court a libel for divorce. That court refused to accept the libel since the alleged cause of divorce occurred outside of Massachusetts and Fiorentino had not resided in the Commonwealth for the two years required by the statute. Fiorentino then filed a petition for mandamus to compel the judges of the Probate Court to accept the libel. After hearing, a single justice of this court, at the request of the parties, reserved and reported the case without decision to the full court.[4]

Carmen Fernandez moved to Massachusetts in September, 1971.[5] In the fall of 1972 she filed a libel for divorce in the Probate Court for Suffolk County. Confronted with the libellant’s challenge to the constitutionality of the two-year residence requirement, the probate judge reported the case to the Appeals Court.[6] The case is before us on direct appellate review. G.L.c. 211A, ß 12, inserted by St. 1972, c. 740, ß 1.

The statutes under challenge operate to define and limit the jurisdiction of the courts of Massachusetts to grant divorces. Old Colony Trust Co. v. Porter, 324 Mass. 581, 586 (1949). General Laws c. 208, ß 4, forbids the granting of a divorce in two situations: where the parties have never lived together as husband and wife in Massachusetts, and where the legal cause of the divorce occurred outside of this Commonwealth. There are certain exceptions to these basic rules. Even if the parties had not lived here as husband and wife, a divorce may be decreed if the cause occurred within Massachusetts and the libellant lives here when the libel is filed. ß 5. Also, if the cause occurred elsewhere a divorce may be decreed by Massachusetts courts if the parties had lived together as husband and wife in Massachusetts prior to the occurrence of such cause and one of them lived here at the time the cause occurred. ß 4. Finally, in the same situation of a cause occurring elsewhere, a divorce may be decreed if the libellant has lived in Massachusetts for two years before filing the libel. ß 5. Thus, the two-year residence requirement applies only to those divorce libellants who rely on a cause of action which occurred outside of Massachusetts and which did not affect a marriage previously existing in the Commonwealth.

These statutory restrictions on the divorce powers of Massachusetts courts were presumably intended to prevent the bringing of migratory causes of action in Massachusetts courts and to ensure the validity (i.e., full faith and credit in other jurisdictions) of the decrees of Massachusetts courts against collateral attack by limiting proceedings in divorce actions to situations where the Commonwealth has some substantial connection with the dispute being adjudicated. In the typical case these interests are served by the court’s refusing to proceed unless it has personal jurisdiction over both parties to a dispute (in personam jurisdiction) or it has jurisdiction over the thing, or res, that is the subject of a dispute (in rem jurisdiction). Jurisdiction over causes of action for divorce, however, is strictly neither in personam nor in rem. Williams v. North Carolina, 317 U.S. 287, 297 (1942). “Under our system of law, judicial power to grant a divorce ó jurisdiction, strictly speaking ó is founded on domicil.” Williams v. North Carolina, 325 U.S. 226, 229 (1945). See Shaw v. Shaw, 98 Mass. 158 (1867). Thus, it is established that State courts may exercise divorce jurisdiction based solely on the domicil[7] of the libellant even if the libellee neither appears nor is personally served and even though the parties never resided as husband and wife in the forum State. Williams v. North Carolina, 317 U.S. 287, 298-299 (1942). In such a case, however, it is essential that the libellant’s claim to domiciliary status in the forum State be a legitimate one; otherwise, the court’s power to dissolve the marriage is a mere chimera vulnerable to collateral attack in another jurisdiction. Williams v. North Carolina, 325 U.S. 226 (1945). Andrews v. Andrews, 188 U.S. 14 (1903). Cohen v. Cohen, 319 Mass. 31 (1946).

It may be because jurisdiction founded solely on the libellant’s domicil was thought to be so fragile that the Legislature has imposed the two-year residence requirement of G.L.c. 208, ß 5. In effect, however, ß 5 creates an irrebuttable presumption that a divorce libellant is not a domiciliary of Massachusetts until he or she has resided in the Commonwealth for at least two years. Such a presumption surely serves to protect the divorce courts of Massachusetts from actions based on fraudulent claims of domicil by temporary residents. The present libellants insist, however, that it also operates to deprive domiciliaries such as themselves,[8] namely, those who have resided here for less than two years, of their constitutional rights of due process and equal protection.[9] We hold that the two-year residence requirement does work a deprivation of equal protection in these cases and, therefore, that it was unconstitutional to bar these libellants from prosecuting their divorce libels solely on the ground that they were not two-year residents of Massachusetts.

The two-year residence requirement operates to classify Massachusetts domiciliaries into two groups based on length of residence and to treat those groups quite differently with respect to access to divorce courts. Under recent United States Supreme Court decisions classifications based on length of residence are constitutionally suspect because they burden the “fundamental” right to travel by penalizing new arrivals in a State relative to longer-term residents of that State. Shapiro v. Thompson, 394 U.S. 618 (1969). Dunn v. Blumstein, 405 U.S. 330 (1972). Memorial Hosp. v. Maricopa County, 415 U.S. 250 (1974). In the Dunn case the Supreme Court overturned Tennessee statutes which barred persons from registering to vote (and thus from voting) who had not resided in the State for at least one year and in the county of registration for three months. Because those durational residence requirements both penalized persons on the basis of recent travel and also denied a fundamental political right (voting), the court held that they could not stand absent a showing by the State that it had a compelling interest in imposing the requirement. Likewise, in the cases before us the two-year residence requirement penalizes persons merely because they have recently traveled interstate.[10] In addition, the benefit withheld by the requirement (i.e., access to the divorce courts) is one that merits constitutional protection. See Boddie v. Connecticut, 401 U.S. 371 (1971). Therefore, as was said in the Dunn case, “[W]hether we look to the benefit withheld by the classification… or the basis for the classification … we conclude that the State must show a substantial and compelling reason for imposing durational residence requirements.” 405 U.S. at 335 (1972).

We believe that the Commonwealth does have a substantial and compelling reason in these cases. Certainly its interest in protecting its courts from fraudulent invocations of their jurisdiction is sufficient reason for imposing requirements aimed at ensuring that divorce libels are brought only by legitimate domiciliaries of Massachusetts. This is not enough, however, to save the two-year residence requirement. The Equal Protection Clause of the Fourteenth Amendment to the Constitution of the United States requires not only that there be a compelling State interest for a classification based on length of residence, but also that the classification be precisely tailored to achieve its legitimate objective and that there be no alternative, less burdensome, means to achieve the State’s goal. Dunn v. Blumstein, 405 U.S. at 343 (1972). Memorial Hosp. v. Maricopa County, 415 U.S. 250, 268 (1974). The issue, therefore, comes down to this: Is the durational residence requirement the least drastic means by which the jurisdiction of the Commonwealth’s divorce courts may be protected from fraud? We hold that it is not.

In the Dunn case, the Supreme Court held that the State’s interest in ensuring “the `purity of the ballot box'” was indeed compelling, 405 U.S. at 345, but the means chosen to fulfil that interest were imprecise and unnecessary. The court found that the durational residence requirements were ill-suited to the goal of preventing non-domiciliaries from casting fraudulent votes because a potential registrant could just as easily lie about the length of his residence as he could about his place of domicil. 405 U.S. 346-347 (1972). See also Memorial Hosp. v. Maricopa County, supra, at 268. In addition, the State had provided an alternative means of discovering and preventing fraud through its voter registration system which permitted a case-by-case review of the legitimacy of an individual’s claim of domicil. 405 U.S. 346-348. See also Memorial Hosp. supra, at 268.

We believe the same analysis and reasoning apply here. Clearly, divorce libellants intent on perjuring themselves might just as readily lie about the length of their residence in Massachusetts as about the fact of their domicil. Thus, the two-year requirement could serve to bar only those who tell the truth. And the argument for a case-by-case determination of domicil, without reliance on any arbitrary presumptions of nondomicil based on length of residence, is in one respect even stronger here than in the Dunn case because divorce libels are tried individually by judicial tribunals which are precisely suited to require proper evidence and to make the findings of fact which are necessary to determine domicil. To urge that the judges of our Probate and Superior Courts are incapable of making proper determinations of domicil in divorce cases is to argue that they are incapable of performing the functions of their offices, namely, finding the facts and applying the law. We see no reason to arrive at such a conclusion.

Those who contend otherwise apparently do not agree that judges can make determinations of domicil on a case-by-case basis with sufficient accuracy to protect the Commonwealth’s divorce courts from fraud. They would therefore uphold the durational residence requirement as a necessary evil. In effect, they perceive the problem of fraudulent domicil claims in divorce cases as so serious as to outweigh any harm caused to legitimate domiciliaries who are barred from divorce courts because they do not satisfy the residence requirements. The difficulty with this viewpoint, however, is that the fraud with which they are so concerned is hypothetical only, while the harm to legitimate domiciliaries is actual and immediate.

In the first place, it is highly unlikely, in view of our strict limitation on the substantive grounds for divorce,[11] that this Commonwealth will be a very appealing jurisdiction for migratory divorce seekers even in the absence of a durational residence requirement. Thus, the possibility of a substantial incidence of fraudulent invocations of Massachusetts divorce jurisdiction depends on the purely speculative possibility that the Legislature will some day liberalize the substantive grounds for divorce. In the meantime, of course, legitimate domiciliaries are denied access to divorce courts in order that the Commonwealth may protect those courts from speculative as well as hypothetical fraud.

More importantly, we do not believe that it can be shown that fraudulent claims of domicil can be perpetrated so effectively as to defy detection by judges who act affirmatively to prevent such fraud. In contexts other than divorce we rely on trial court judges’ ability to make valid findings of domicil. See, e.g., Plymouth v. Kingston, 289 Mass. 57 (1935); Rummel v. Peters, 314 Mass. 504 (1943). Although it is likely true that the interest of a person in obtaining a dissolution of his or her marriage may provide greater motivation for perjury in divorce cases than in other situations, it is, nevertheless, also true that determination of domicil remains “mainly a question of fact,” Tuells v. Flint, 283 Mass. 106, 109 (1933), as to which there exists “a myriad of tangible criteria” which are “highly relevant” to that determination. Mon Chi Heung Au v. Lum, 360 F. Supp. 219, 222 (D. Hawaii 1973). Among such criteria length of residence would be one, but only one, relevant consideration. Any evidence indicating an intention to establish roots would be material.[12] See Comment, 30 Md. L. Rev. 367, 380, n. 113 (1970). Given such evidence and observation of the libellant’s demeanor, and backed by the dual sanctions of the perjury statutes[13] and the potential invalidity of a decree issued without jurisdiction, a judge sitting in a divorce case can make a reasonably accurate determination whether the libellant’s claim of Massachusetts domicil is bona fide.[14]

It is worth noting that G.L.c. 208, ßß 4 and 5. themselves contain implicit legislative assertions of confidence in the ability of the judges of Massachusetts divorce courts to determine domicil without having to rely on durational residence requirements. For example, the statutes in several situations require a determination whether the parties ever “lived together as husband and wife in this commonwealth.” This court has consistently construed that provision as requiring a showing that the parties were domiciled here as husband and wife. Winans v. Winans, 205 Mass. 388, 393 (1910). Field v. Field, 236 Mass. 256 (1920). Thus, to determine whether jurisdiction is proper on the ground that the parties lived here as husband and wife, a judge must make a determination of domicil without reliance on a durational residence requirement. Likewise, in so far as the words “lived” in the final clause of ß 4 and “resident” in ß 5 are construed to require domicil rather than mere residence,[15] to that extent the Legislature has declared that judges are capable of making case-by-case determinations of domicil. Such determinations become no more difficult just because domicil of the libellant is the sole connection between Massachusetts and the marriage, as in the cases before us.

It has been suggested that the two-year residence requirement is justified because only after such a period is the Commonwealth’s interest in a resident sufficient to override the injustice which may result to the libellee in a divorce proceeding where only the libellant appears personally. There are two problems with this argument. First, any injustice which may be caused the absent spouse (because of notice by mail or by publication, or lack of means to enter an appearance) will not be mitigated but in fact might be exacerbated by a two-year delay which makes it increasingly unlikely that the nonresident could be located for notice purposes.[16] Second, there is no substantial justification for the assertion that the Commonwealth’s “interest” in a resident somehow increases in proportion to the length of residence. Absent some other valid basis for classification, the Commonwealth’s obligation is to treat all residents alike, regardless of length of residence. Dunn v. Blumstein, 405 U.S. 330 (1972).

Finally, we note that, in addition to its invalidity under the Equal Protection Clause, the two-year residence requirement may well be vulnerable to due process attack on two grounds. First, the Supreme Court has with increasing frequency asserted that “[s]tatutes creating permanent irrebuttable presumptions have long been disfavored under the Due Process Clause of the Fifth and Fourteenth Amendments.” Vlandis v. Kline, 412 U.S. 441, 446 (1973). See Stanley v. Illinois, 405 U.S. 645 (1972); Cleveland Bd. of Educ. v. LaFleur, 414 U.S. 632, 644 (1974). See also Bell v. Burson, 402 U.S. 535 (1971). The presumption against domiciliary status raised by G.L.c. 208, ßß 4 and 5, is certainly “irrebuttable” and is arguably as “permanent” as the presumption struck down in the LaFleur case, supra (teachers’ mandatory maternity leave requirement creates presumption of physical incapacity during final months of pregnancy).[17] Because the statutory presumption now before us infringes on certain constitutionally protected interests of divorce libellants (see discussion below), and because that presumption is “neither “necessarily nor universally true,'” Cleveland Bd. of Educ. v. LaFleur, supra, at 646, it may be that the two-year residence requirement is unconstitutional under the Supreme Court’s due process analysis of irrebuttable presumptions. Cleveland Bd. of Educ. v. LaFleur, supra.

Second, a related but separate due process issue arises from Boddie v. Connecticut, 401 U.S. 371 (1971), in which the Supreme Court held that it is constitutionally impermissible to bar indigents from divorce courts solely on the basis of their inability to pay a filing fee. Emphasizing that “marriage involves interests of basic importance in our society” (id. at 376) and that the mechanism for dissolving marriage is monopolized by the State, the majority asserted that the State’s interest in charging court fees was not sufficient to override the interests of indigents in having access to divorce courts. It therefore held that the filing fee requirement violated the due process rights of indigent divorce libellants. In the cases now before us, the libellants argue that the two-year residence requirement operates to deny them access to divorce courts just as surely as the filing fee requirement struck down in the Boddie case denied such access to indigents. Therefore, conclude the libellants, because the State can put forward no interest in the durational residence requirement substantial enough to override the interest which domiciliaries of less than two years’ residence have in access to the divorce courts, the two-year residence requirement of G.L.c. 208, ßß 4 and 5, is unconstitutional under the Boddie holding.

Regardless of the proper due process analysis, however, we conclude that the two-year residence requirement of G.L.c. 208, ßß 4 and 5, at the very least, is unconstitutional as violative of the Equal Protection Clause. A durational residence requirement is simply too imprecise a means for accomplishing the goal of restricting Massachusetts divorces to legitimate Massachusetts domiciliaries.[18] In addition, there is available the more precise, less burdensome alternative of case-by-case judicial determination of domicil. Given the certainty that the present system bars legitimate domiciliaries access to our divorce courts, we hold that the harm thus caused outweighs any harm which might result from the speculative possibility of fraudulent claims to domicil. The libellants in the cases now before us should be permitted to prove the legitimacy of their assertions of Massachusetts domicil without having to wait out an arbitrary two-year period as required by the offending statute.

Case No. S-7837 is remanded to the Probate Court for further proceedings. The petition for writ of mandamus in case No. S-15,276 is allowed.

So ordered.

REARDON, J. (dissenting, with whom Quirico and Braucher, JJ., join).

The Fernandez case became moot in September, 1973. The Fiorentino case will become moot on April 1, 1974. Nevertheless, as the cases were properly brought before the two-year residence requirement expired, we would ordinarily agree with the majority that we should declare the rights of the parties. See Karchmar v. Worcester, 364 Mass. 124, 136 (1973). In this case, however, we are adjudicating a difficult and important Federal constitutional question on which other courts have divided and which the Supreme Court of the United States will doubtless soon resolve (see fn. 2 to the majority opinion). In such a situation we should not rush to make a constitutional decision which cannot finally resolve the question. See First Natl. Bank v. Attorney Gen. 362 Mass. 570, 596 (1972) (concurring opinion of Quirico, J.).

In the Fernandez case the judge did not decide the question of jurisdiction, but, “acting under the provisions of G.L.c. 215, ß 13,” reserved and reported the entire case. The power to reserve and report under that section is limited to cases in which interlocutory decrees have been made and cases that have been heard for final determination. It does not appear that this case is one in either class, and it is therefore not properly here on report. Second Bank-State St. Trust Co. v. Linsley, 341 Mass. 113, 116 (1960). Compare Bennett v. Florence, 347 Mass. 707, 709 (1964).

In the Fiorentino case the Probate Court refused to accept the libel for filing, and the libellant filed a petition for mandamus to compel acceptance of the libel. A similar situation arose in Peace v. Peace, 362 Mass. 536, 538 (1972), where we said that “such an ex parte presentation without the issuance of process for service on the libellee and an attempt to bring him within the jurisdiction of the Probate Court is not a satisfactory basis upon which to require our decision of such an important constitutional question.” On the authority of that case, the present case should be returned to the Probate Court with a direction to allow the filing of the libel. After service of appropriate process on the libellee, the case should proceed in due course. Meanwhile, we should make no decision and intimate no opinion on the constitutionality of G.L.c. 208, ßß 4 and 5.

Since this court nevertheless decides the constitutional question, we express our views on it. We believe that it is constitutionally permissible for the Legislature to determine that the durational residence requirement imposed on the libellants in these cases by G.L.c. 208, ßß 4 and 5, is necessary to protect State interests of the most compelling nature. Therefore we must respectfully dissent from the decision and opinion of the court.

The majority opinion correctly delineates the equal protection issues raised by these cases.[1] There is no question that this residence requirement does place some burden on some new residents of the Commonwealth by requiring them to wait for a specified period before they may properly bring a libel to dissolve a marriage which has no connection with Massachusetts beyond the asserted domicil of the libellant. Furthermore, because the burden imposed impinges on the critical personal right of access to divorce courts, Boddie v. Connecticut, 401 U.S. 371, 376 (1971), it may “penalize” the exercise of the right to interstate travel. See Memorial Hosp. v. Maricopa County, 415 U.S. 250, 258-259 (1974). But in our view this does not end the inquiry. The Supreme Court has clearly held that durational residence requirements need not be offensive to the Constitution in every case. Shapiro v. Thompson, 394 U.S. 618, 638, n. 21 (1969). Starns v. Malkerson, 401 U.S. 985 (1971), affg. without opinion 326 F. Supp. 234 (D. Miss. 1970). Vlandis v. Kline, 412 U.S. 441, 452-453, n. 9 (1973). Sturgis v. Washington, 414 U.S. 1057 (1973), affg. without opinion 368 F. Supp. 38 (W.D. Wash. 1973). Memorial Hosp. v. Maricopa County, 415 U.S. 250, 256 (1974). Although it is not clear in these cases whether or not the court found a significant “penalty” inherent in the restriction, Memorial Hosp. v. Maricopa County, supra at 257-259, the court has made it plain that even such a penalty may be permissible if necessary to promote a compelling State interest. See, e.g., id. at 258; Shapiro v. Thompson, supra, at 634; Dunn v. Blumstein, 405 U.S. 330, 341-343 (1972).

It is true that Chief Justice Burger, commenting on this “strict scrutiny” standard of review in his dissent in Dunn v. Blumstein, supra, noted: “So far as I am aware, no state law has ever satisfied this seemingly insurmountable standard, and I doubt one ever will, for it demands nothing less than perfection.” 405 U.S. at 363-364. It has been suggested that the use of the “active” standard of review is a mere camouflage for striking down classifications which are unsatisfactory for other unspoken reasons, see Note, 58 Va. L. Rev. 1489, 1495-1496 (1972), and that this standard provided review which was “`strict’ in theory and fatal in fact.” Gunther, Foreword: In Search of Evolving Doctrine on a Changing Court: A Model for a Newer Equal Protection, 86 Harv. L. Rev. 1, 8 (1972). We cannot take so cynical a view of the language of the Supreme Court. If that court had determined that certain classifications were per se unconstitutional irrespective of claimed State justification, language was at its disposal to say just that. It has not done so. It has stated only that such classifications cannot be sustained “unless the State can demonstrate that such laws are `necessary’ to promote a compelling governmental interest.'” Dunn v. Blumstein, supra, at 342. Taking the court at its word and finding such a necessity here, we would uphold the statute. See Shiffman v. Askew, 359 F. Supp. 1225, 1229 (M.D. Fla. 1973).

As the majority recognize, the challenged statutes are jurisdictional in nature. Old Colony Trust Co. v. Porter, 324 Mass. 581, 586 (1949). Thus the interest they serve is the interest of a State in putting some limits on the kinds of causes and parties which come before its courts. The ability so to limit causes and parties is not only inherent in the concept of sovereignty of a State but is an essential element of that comity which makes our Federal union possible. That Federal system relies upon State court restraint with respect to causes having only minimal ties to the respective States as well as upon that full faith and credit afforded to the judgments of courts of sister States mandated by art. 4, ß 1, of the United States Constitution. See Hazard, A General Theory of State-Court Jurisdiction, 1965 Sup. Ct. Rev. 241, 246-247. Compare Hanover v. Turner, 14 Mass. 227 (1817); Galpin v. Page, 18 Wall. 350 (1873); Story, Commentaries on the Conflict of Laws, ß 20. II (8th ed. 1883). The failure of States so to confine their jurisdiction would play havoc with the delicate balance of comity. See von Mehren & Trautman, Jurisdiction to Adjudicate: A Suggested Analysis, 79 Harv. L. Rev. 1121, 1126 (1966). The interest of the Commonwealth in maintaining that balance may certainly be deemed not only “overriding” or “compelling” but indeed fundamental to the very nature of its existence as a component of the Federal union.

The majority further recognize that the interest of the State in maintaining the jurisdictional integrity of its divorce courts presents a particularly difficult task because, unlike jurisdiction in other actions, divorce jurisdiction may constitutionally be founded on the mere domicil of one party. It is appropriate for States to supplement this slender jurisdictional base with a strict and objective test such as that provided by G.L.c. 208, ßß 4 and 5. These sections effectively reinforce the critical State objective of restricting the Massachusetts Probate Courts to divorce by bona fide domiciliaries. See Shiffman v. Askew, supra, at 1231-1232. By so doing, they manifest the Commonwealth’s concern for comity by refusing to adjudicate matters which are more properly left to the law of other jurisdictions. They also make it more certain that Massachusetts decrees will be respected in other jurisdictions which could refuse them effect on the basis of improper domicil. Williams v. North Carolina, 325 U.S. 226, 239 (1945).

Although they acknowledge the compelling character of this interest, and also concede that the statute struck down today protects that interest, the majority hold that there exists a less restrictive means of achieving the interest, namely, a case by case examination of each potential libellant to determine whether he or she possesses the requisite domiciliary intent to establish jurisdiction. The majority thus adopt the reasoning of the Supreme Court in striking down durational residence requirements for voting in Dunn v. Blumstein, supra. As in the opinion in the Dunn case, the majority cite a list of other objective criteria for determining domicil without resort to durational residence requirements. This method may well be effective as a way of detecting bona fide residents for voting purposes. But the mechanical application of that approach to the problem of determining domicil for divorce jurisdiction reveals an unwillingness to come to grips with the special characteristics of the litigants and proceedings involved in divorce.

The majority have, in fact, completely ignored the universally unhappy experience of divorce courts when relying on anything other than strict objective tests. Perhaps no other sector of the judicial process has been so fraught with sham and deception as that dispensing divorce. The sense of urgency which parties bring to a divorce action, when it comes up against the barriers erected by State law in pursuance of important policies of either a jurisdictional or a substantive nature, has frequently overcome what might normally be an extreme reluctance to lie to a court. Such fraud has often been used to evade stringent grounds for divorce. In the period when the sole ground for divorce in the State of New York was adultery, one State commission found “widespread fraud, perjury, collusion and connivance in matrimonial actions of every type.” Note, New York’s New Divorce Law: Beyond the Sixth Commandment, 5 Col. J. of Law and Social Problems, 1, 2, n. 12 (No. 2, 1969). See Hawke, Divorce Procedure: A Fraud on Children, Spouses and Society, 32 Texas Bar J. 163 (1969); Westman, The Myth of the Uncontested Divorce, 41 Fla. Bar J. 304 (1967). Perjury “has sadly become a matter of course in many states” when it comes to the needed allegations of domicil. Ehrenzweig & Louisell, Jurisdiction State and Federal In A Nutshell, ß 10 (3d ed. 1973). Not only is perjury a danger in obtaining jurisdiction for divorce, but parties are willing (in a way they would be unlikely to be in the case of voter fraud) to acquire the indicia of domicil so heavily relied upon by the majority as the subject of an individualized inquiry. The acquisition of a divorce decree is regarded as such an unusual occurrence in individual lives and of such paramount importance that it is understandable why many will “manufacture” a domicil which they have no intention of retaining. See Ehrenzweig, Conflict of Laws, ß 72 (1959). Commenting on this phenomenon in the State of Nevada, one observer noted: “All the while they contemplated returning to their home states immediately after their divorces were secured, yet they all swore falsely that they intended to make Nevada their permanent home, having been warned by local counsel that, unless they did so, they would be out of court. On advice of counsel they also took steps which would be accepted by the Nevada courts as corroborating their sworn statement but were actually nothing more than sham and camouflage.” Lorenzen, Extraterritorial Divorce ó Williams v. North Carolina II, 54 Yale L.J. 799, 801 (1945), quoted in Whitehead v. Whitehead, 53 Hawaii 302, 308, fn. 3 (1972). It seems apparent that the extensive litany of factors bearing on domicil which the majority would have probate judges investigate is utterly inadequate to deter or discover the possible fraud of those who perceive so much at stake in a unique episode in their lives.

The issue is not as stated by the majority whether “the problem of fraudulent domicil claims in divorce cases … [is] so serious as to outweigh any harm caused to legitimate domiciliaries who are barred from divorce courts.” Having admitted that the State interest in the jurisdiction of its courts is a compelling one, the only question is whether the Legislature may properly determine that no test of domicil other than a durational residence requirement adequately protects that critical State interest in the context of divorce. If, as in Massachusetts, it has concluded that no other tests can be effective, it is not for this court to balance the relative injuries incurred by individuals in preserving that interest.

Nor is the question here one of the competence of judges in “performing the functions of their offices.” It is not the qualities of the judges but the incentives of the parties which are relevant and which set the area of divorce apart from other court proceedings. Even the most perspicacious judge cannot separate the new but legitimate domiciliary from the dissembling transient if the latter is both willing and able to acquire the evidence which the majority indicate would be sufficient to satisfy an individualized test.[2]

It is certainly true, as the majority point out, that the parties may lie about two years of residence as well as about domiciliary intent. What the majority do not consider, however, is that the risk of such deceit is considerably reduced in the former cases. There is no doubt a greater reluctance to swear falsely to facts which are already false than to some uncertain future intent or state of mind. Moreover, in doubtful cases the fact of continuous residence for the required period may more easily be proved or disproved than the domiciliary intent evidenced only by sworn statements and easily acquired “objective” indicia. By requiring two years of residence in addition, the Legislature has adopted a test which most effectively secures the Commonwealth’s interest in granting divorces only to bona fide domiciliaries. The statute does not, and cannot, operate with complete perfection, including all legitimate domiciliaries and excluding only those seeking migratory divorces. But no classification can sustain that burden, and this should not automatically disqualify it under the strict constitutional test applied. That test does “not have the precision of mathematical formulas” but “emphasize[s] a matter of degree.” Dunn v. Blumstein, 405 U.S. 330, 342-343 (1972). It is for the Legislature to determine the least restrictive means which are reasonably adapted to the protection of the vital State interest involved.

We also dispute the majority’s contention that stringent safeguards on the jurisdiction of our Probate Courts are unnecessary because of the strict limitations on the substantive grounds for divorce in Massachusetts. “Cruel and abusive treatment” may be established by mere words, and the granting of a divorce is not discretionary. Brown v. Brown, 323 Mass. 332, 334 (1948). La Raia v. La Raia, 329 Mass. 92 (1952). Even if the present substantive grounds are regarded as strictly limited, they can and do experience change.[3] Many States are currently undergoing a reexamination of their substantive divorce laws and a number have adopted the concept of “no-fault” divorce. See, e.g., Fla. Sts. ß 61.052 added by Laws 1971, c. 71-241, ß 7; Iowa Code (1966) ß 598.5, added by Acts of 1970, c. 1266, ß 6. The majority would have Massachusetts protect its compelling interest in assuring the proper domicil of parties in divorce actions by assuming that its substantive law of divorce cannot be liberalized. It is competent, however, for the Legislature to determine that the current mechanism of requiring continued residence in specified cases is far more appropriate and that it is compatible with any grounds for divorce the Legislature may hereafter formulate.

In short it seems clear that the danger to the jurisdictional integrity of Massachusetts divorce courts is a significant one. By hedging about the requirement of domicil in those cases where that alone provides the nexus between the Commonwealth and the marriage which is the subject of the libel, the Legislature has chosen a means carefully suited to the end sought. The history and experience of divorce courts show clearly that nothing less will do. The loss of that protection opens the door to possible fraud and abuse which mock the very judicial process which is being invoked.

The present application of the statute also serves the distinct State interest in assuring justice to litigants. These cases suggest the hardship and possible injustice which may result from an exercise of divorce jurisdiction on the basis of domicil without other safeguards. In the Fiorentino case the libellee wife continues to reside in New York with the two children of the parties, and the husband has been ordered to pay $50 a week for their support. The record discloses no personal service on the wife, no actual notice to her, no appearance on her behalf in the present proceeding, and only notice “by certified mail return receipt requested” in the prior separate support proceeding. In the Fernandez case the record shows that the libellee husband, a resident of Puerto Rico, was served by “publication and registered mail,” and a return receipt indicates that he had actual notice. In neither case is there any showing that the libellee has the means to participate in judicial proceedings in Massachusetts.

It is constitutional for our courts to exercise divorce jurisdiction without jurisdiction of the person of the libellee, even though the libellee has never had any significant contract with the Commonwealth of Massachusetts. In such cases a finding of domicil could be relitigated in another forum, and our courts have no power to bind the absent spouse personally by an adjudication with respect to support or property rights. Our Uniform Reciprocal Enforcement of Support Act provides an effective procedure for an absent wife to enforce her right to support against her husband in Massachusetts. G.L.c. 273A. See M—- v. W—-, 352 Mass. 704, 708 (1967). But an effective divorce decree, obtained ex parte, might terminate her right to support. See Madden v. Madden, 359 Mass. 356, 365 (1971), and cases cited.

In these circumstances we do not think the Constitution compels the Commonwealth to exercise its jurisdiction to grant ex parte divorces. The interest of the Commonwealth in assuring justice to the absent spouse is a compelling one. The delay of two years permits the parties to work out their difficulties either by reconciliation or by a divorce in the State where the absent spouse remains. As time passes, the interest of the absent spouse in the affairs of the refugee in Massachusetts is likely to diminish, and the former home in New York or Puerto Rico may come to seem less appropriate as a forum for litigation. Meanwhile, the ties of the new resident to Massachusetts are likely to deepen and grow stronger, and the chances that a finding of domicil in Massachusetts can be contested elsewhere are progressively reduced. When the balance swings sufficiently to justify the risk of injustice to the absent spouse is peculiarly appropriate for legislative rather than judicial determination. It cannot be said that a two-year period is excessive, much less that it is irrational or invidious. Compare Boddie v. Connecticut, 401 U.S. 371, 388-389 (1971) (concurring opinion of Brennan, J.).

The Supreme Court of the United States has not held that durational residence requirements for divorce constitute a “penalty” in the constitutional sense in any circumstances, and certainly has not held them unconstitutional in the present circumstances. Durational residence requirements have been held to be “penalties” in cases involving eligibility for welfare, voting, and medical care, but they have been held valid with respect to tuition charges at State universities. The difference is that educational tuition is neither “a fundamental political right” nor “a basic necessity of life.” Memorial Hosp. v. Maricopa County, 415 U.S. 250, 259 (1974). The same may be said of ex parte divorce.

[1] ß 4 “A divorce shall not, except as provided in the following section, be decreed if the parties have never lived together as husband and wife in this commonwealth; nor for a cause which occurred in another jurisdiction, unless before such cause occurred the parties had lived together as husband and wife in this commonwealth, and one of them lived in this commonwealth at the time when the cause occurred.”

ß 5, as appearing in St. 1969, c. 162. “If the libellant has lived in this commonwealth for two years last preceding the filing of the libel if the cause occurred without the commonwealth, or if the libellant is a resident of the commonwealth at the time of the filing of the libel and the cause occurred within the commonwealth, a divorce may be decreed for any cause allowed by law, unless it appears that the libellant has removed into this commonwealth for the purpose of obtaining a divorce.”

[2] There is no consensus among the courts which have recently passed upon the constitutionality of durational residence requirements in divorce cases. Compare Wymelenberg v. Syman, 328 F. Supp. 1353 (E.D. Wis. 1971); Mon Chi Heung Au v. Lum, 360 F. Supp. 219 (D. Hawaii, 1973); Larsen v. Gallogly, 361 F. Supp. 305 (D.R.I. 1973) (all striking down durational residence requirements), with Shiffman v. Askew, 359 F. Supp. 1225 (M.D. Fla. 1973); Sosna v. Iowa, 360 F. Supp. 1182 (N.D. Iowa, 1973), review granted 415 U.S. 911 (1974); Whitehead v. Whitehead, 53 Hawaii 302 (1972); Porter v. Porter, 112 N.H. 403 (1972); Place v. Place, 129 Vt. 326 (1971) (all upholding durational residence requirements).

[3] For convenience, we refer to petitioner Fiorentino as a “libellant.”

[4] In the mandamus proceeding the respondent Probate Court judges have been represented by a special assistant attorney general. Both in a brief and during oral argument the special assistant attorney general has thoroughly and effectively advanced the arguments for upholding the constitutionality of the two-year residence requirement. Thus, the reason for our reluctance to decide the constitutional issue in Peace v. Peace, 362 Mass. 536 (1972), i.e., that the case had been argued on an ex parte basis, is not present here.

[5] Although Fernandez has now fulfilled the residence requirement, and presumably may now file a proper libel under the statute, we nevertheless shall proceed to a decision because the companion case is still at issue and because of the importance and recurring nature of the issue presented. See Dunn v. Blumstein, 405 U.S. 330, 333, n. 2 (1972).

[6] The judge reported the case pursuant to G.L.c. 215, ß 13. In the dissenting opinion, the Justices suggest that this case is not properly before us because ß 13 does not authorize the reservation and report of a case in which there has been no interlocutory decree and which has not been heard for final determination. We do not agree that this case has not properly been reported. By reporting the case the judge has decreed, in effect if not precisely in form, that he has no jurisdiction to hear the case. The question of Probate Court jurisdiction is one which we have previously considered on report under ß 13 where the report is preceded by a decree dismissing the libel for lack of jurisdiction. Bennett v. Florence, 347 Mass. 707 (1964). There is no practical reason to distinguish this case from the Bennett case. In any event, the jurisdictional issue is one of public importance. “[I]t has been fully argued to us, and even if we had found it necessary to dismiss the report as not properly before us, we should neverthless have felt warranted in expressing our views.” Commonwealth v. Haddad, 364 Mass. 795 (1974), and cases cited.

[7] The meaning of “domicil,” as opposed to “residence,” has often been explained. “In general it is said to be the place of one’s actual residence with intention to remain permanently or for an indefinite time and without any certain purpose to return to a former place of abode.” Tuells v. Flint, 283 Mass. 106, 109 (1933). “[T]he acquisition of a domicil does not depend on the length of the residence.” Winans v. Winans, 205 Mass. 388, 392 (1910).

[8] In the present postures of these cases, we accept the libellants’ assertions that they are bona fide domiciliaries of Massachusetts (assertions which must eventually be established at trial).

[9] The libellants do not dispute the Commonwealth’s interest in restricting divorce jurisdiction to legitimate domiciliaries. Their challenge is directed only at the durational residence requirement.

[10] No showing of actual deterrence of travel need be made. It is enough to show that the classification in question penalizes the exercise of the right. Dunn v. Blumstein, 405 U.S. at 339-340 (1972). Although not all waiting periods are “penalties,” see Shapiro v. Thompson, 394 U.S. at 638, n. 21 (1969), we believe, in light of Boddie v. Connecticut, 401 U.S. 371 (1971), that denial of access to divorce courts imposed by the two-year residence requirement amounts to a penalty in the constitutional sense. See Memorial Hosp. v. Maricopa County, 415 U.S. at 258-260 (1974). Thus, the cases before us are distinguished from the cases in which the Supreme Court declined to strike down State statutes requiring one year of residence as a condition to lower tuition at State institutions of higher education. Starns v. Malkerson, 401 U.S. 985 (1971), affg. 326 F. Supp. 234 (D. Minn. 1970). Sturgis v. Washington, 414 U.S. 1057 (1973). affg. 368 F. Supp. 38 (W.D. Wash. 1973). Compare Vlandis v. Kline, 412 U.S. 441 (1973). A denial of lower tuition does not amount to a constitutional penalty. Memorial Hosp. v. Maricopa County, 415 U.S. at 260, n. 15 (1974).

[11] See G.L.c. 208, ßß 1 and 2.

[12] Factors which might be considered would be: whether the libellant has a Massachusetts driver’s license and automobile registration; whether he or she has purchased a home or has leased an apartment in the Commonwealth; the term of any apartment lease; whether any children have been brought to live in Massachusetts; whether personal property, including household goods, has been brought here; whether permanent employment has been obtained in Massachusetts; whether there is evidence of abandonment of previous domicil, e.g., cancellation of bank accounts, leases, memberships, and so forth, sale of property, and issuance of change of address notices. This list is not meant to exhaust the possible factors to be considered.

[13] See G.L.c. 268, ßß 1-4.

[14] In addition, the burden of proving his or her domicil will be on the libellant. Mellon Natl. Bank & Trust Co. v. Commissioner of Corps. & Taxn. 327 Mass. 631, 638 (1951).

[15] It is not clear whether the words “lived” and “resident” in ßß 4 and 5 are in each instance to be construed as meaning “domicil” and “domiciliary.” Whenever the question has arisen, this court consistently has held, or at least assumed, that domicil is required. Shaw v. Shaw, 98 Mass. 158, 159 (1867). Winans v. Winans, supra. Labonte v. Labonte, 210 Mass. 319, 320 (1911). Field v. Field, supra. See Sampson v. Sampson, 223 Mass. 451, 461 (1916); Atwood v. Atwood, 297 Mass. 229, 232 (1937). Of course, if the use of “lived” in ß 5 means “domiciled,” see Labonte v. Labonte, supra, then the justification for the two-year residence requirement evaporates altogether. If the requirement is for two years of domicil, then the two-year waiting period is irrelevant as evidence of domicil since domicil must be proved as of the beginning of the period.

[16] The problem is exemplified by the Fernandez case. The Fernandezes were married and lived together in New York. According to the record, the libellee husband now resides in Puerto Rico. It would presumably be just as difficult for him to appear in a divorce proceeding in New York as in one in Massachusetts. In addition, there is no reason to believe that it would become any less difficult for him after two years have passed.

We should emphasize the necessity of strict compliance with the statute and court rules which stipulate notice requirements in divorce proceedings (G.L.c. 208, ß 8, Rules 12 and 14 of the Superior Court [1954] and Rule 41 of the Probate Court [1959]). No divorce decree can have legal effect when it is based on notice to an absent libellee which is inadequate either because it fails to comply with statute or court rule or because it does not meet the requirements of due process. Corkum v. Clark, 263 Mass. 378 (1928). Williams v. North Carolina, 317 U.S. 287, 299 (1942). Actual notice is not required, Atherton v. Atherton, 181 U.S. 155, 172-173 (1901), but the notice which is given must be such that it is “reasonably calculated, under all the circumstances, to apprise interested parties of the pendency of the action.” Mullane v. Central Hanover Bank & Trust Co. 339 U.S. 306, 314 (1950). Judges presiding in ex parte divorce actions should freely exercise the power given them by statute and court rule to ensure that, whenever possible, actual notice is given to absent libellees.

[17] Permanence is, of course, a matter of degree. The presumption at issue in the LaFleur decision, supra, was “permanent” in the sense that the teacher could never return to work while she was pregnant. The practical result of such a “permanent” exclusion, however, is to keep the teacher off the job only for a period of months. The presumption is only as permanent as the pregnancy.

[18] In striking down the two-year residence requirement we, of course, intend no suggestion as to the validity of other requirements of our divorce statutes which involve time delays. See, e.g., G.L.c. 208, ß 8B (husband wife must have lived apart for thirty days before a libel for divorce may be filed). In Dunn v. Blumstein, 405 U.S. 330 (1972), the Supreme Court struck down the durational residence requirements but let stand the State’s provision for closing the voter registration books thirty days prior to election day. Even though that provision barred some new residents (i.e., those who arrived within thirty days of the election) from voting, the court upheld it as administratively necessary to give the State an opportunity to investigate and determine the residence of voter registrants. 405 U.S. at 348. See Marston v. Lewis, 410 U.S. 679 (1973). In the present context such a period between filing the libel and proceeding on the merits might be constitutionally justifiable to permit investigation of the libellant’s claim of domicil. It would not be justifiable as a durational residence requirement for the purpose of proving domicil.

[1] Although the majority decline to rest their opinion upon their interpretation of the applicability of the due process clause to this case, their analysis requires at least a brief comment. It is true that a recent flurry of United States Supreme Court cases has held State statutes unconstitutional because they create “irrebuttable presumptions.” Stanley v. Illinois, 405 U.S. 645, 656-658 (1972). Vlandis v. Kline, 412 U.S. 441, 446 (1973). Cleveland Bd. of Educ. v. LaFleur, 414 U.S. 632, 647-648 (1974). The wisdom of and the exact limits to which the reasoning of these cases may extend is a matter which only the passage of time will reveal. See Cleveland Bd. of Educ. v. LaFleur, supra, at 658-659 (Rehnquist, J., dissenting). The Vlandis case does make clear, however, that not all residence requirement are unacceptable. The majority in that case noted with approval (at 452-453, n. 9) Starns v. Malkerson, 401 U.S. 985 (1971), which upheld a one-year domicil requirement for a lower resident tuition rate at State universities. The distinction drawn there was between presumptions which were permanent and those which were temporary. The present case, of course, is a temporary presumption since continued residence for two years eliminates the presumption. Contrary to the view taken by the majority, the LaFleur case does not dictate a different rule. The presumption examined there did not concern the continuance in time of any status but rather the physical capacity of pregnant women to perform certain employment duties. A reasonable “temporary” presumption about duration of residence may be acceptable while a faulty presumption about pregnancy will offend even though the pregnancy is temporary.

The same distinction with respect to permanence also applies to the majority’s citation of Boddie v. Connecticut, 401 U.S. 371 (1971). The barrier to divorce actions by indigents in that case was a potentially permanent one. The restraint imposed on potential libellants who are subject to the residence requirement of G.L.c. 208, ßß 4 and 5, is by its very nature temporary.

[2] The majority contend that the statutes have already committed individualized determinations of domicil to probate judges by virtue of the requirement in certain cases that the parties have lived together as husband and wife in this Commonwealth which has been interpreted as requiring a previous domicil. That requirement, however, refers to a domicil established before any cause for divorce arose. Thus the incentives to fraud discussed in the text are extremely unlikely to exist. It is considerably easier, of course, to determine the legitimacy of domicil as an historical fact than as a present intention.

The majority opinion, itself, is largely addressed to the propriety of using residence as a test of domicil. Thus the majority do not seriously press their suggestion that the two-year residence requirement of G.L.c. 208, ß 5, really requires two years of continuous domicil. Even if such were the meaning, and probate judges were required to examine the question of domicil during that period without the benefit of a presumption drawn from residence, they would still have the advantage of an historical rather than a psychological investigation. The following two years of actual continued residence (a necessary component of domicil) would be of significant help in determining whether domicil really existed at the beginning of that period.

[3] The requirement for the length of time to make desertion an adequate ground for divorce was liberalized in 1967, reducing the requisite period from three to two years. St. 1967, c. 585, ßß 1 and 2.