OPINION
{1} This is another nationwide class action involving modal premiums. The district court certified a class potentially numbering 700,000 “holders of Old Line insurance policies who purchased their policies at any time between January 1, 1980, and the date of class certification and who have paid modal premiums ... to Old Line.” In line with our opinion in Berry v. Federal Kemper Assurance Co., 2004-NMCA-116, 136 N.M. 454, 99 P.3d 1166 (No. 23,186) (July 23, 2004), we affirm in part, reverse in part, and remand for further consideration.
FACTS AND PROCEDURES
{2} The pleadings in this case follow a familiar pattern. Plaintiff Lisa Enfield alleges that, because she pays quarterly, she is being charged a higher premium by Defendant Old Line Life Insurance Company (Old Line) than is provided for in her policy. Plaintiff Enfield’s policy provides for a quarterly “Billing Frequency” and a “Mode Premium” of $179.14. Plaintiff contrasts these provisions with the “Policy Specifications” page which reflects an annual premium of $676 and a general statement that “premiums other than annual are a percentage of the annual premium.” Plaintiff also relies on premium tables that reflect “Maximum Annual Life Insurance Premium” of $676 for the first twenty years of the policy. Plaintiff also alleges that “Old Line knowingly failed to state or disclose material facts regarding ... payment options under the policy.” Specifically, Plaintiff alleges Old Line did not disclose the dollar difference between the various payment modes available, and, further, did not translate the dollar differential into an “effective annual percentage [or] interest rate.”
{3} Relying on these basic factual allegations, Plaintiff asserted six causes of action: (1) breach of contract (Count I); (2) failure to disclose material facts (Count II); (3) breach of the implied covenant of good faith and fair dealing (Count III); (4) violation of the Unfair Practices Act, NMSA 1978, §§ 57-12-1 to -22 (1978, as amended through 1999) (Count TV); (5) a request for injunctive relief *400(Count V); and (6) a request for declaratory relief (Count VI).
{4} Plaintiff requested nationwide certification for Counts I, II, and III, and a fifteen-state subclass for the Unfair Practices Act claim. After briefing and an evidentiary hearing, the district court refused to certify the subclass but granted the rest of the request. The parties did not file any dispositive motions before dealing with the class issues. Thus, the district court has not considered the merits of any of the claims.
{5} At the certification hearing, the parties stipulated to all of the documentary evidence, except Defendant’s Exhibit CH, which the district court admitted after argument. Exhibit CH is a copy of the file created by Plaintiffs insurance agent during the sale process. The documentary record parallels the pattern found and discussed in our opinion in Berry. The term life policy forms used by Old Line during the class period were essentially identical. The application forms were essentially identical for the class period. The agency agreements and independent marketing organization agreements through which Old Line allowed sale of its policies were essentially uniform throughout the class period. These form agreements all purport to limit the sales person’s ability to modify policies and prohibit the conveyance of any material about Old Line’s policies without its permission.
{6} None of the documentary material designed to reach policy holders contains any information describing Old Line’s modal premium factors and none provides any explicit statement concerning the dollar cost differential inherent in paying more than once a year. Old Line agrees it has never translated the cost differential into an interest rate or APR figure, though it asserts it does not do so because there is no loan involved in the premium structure. None of Old Line’s documentary material specifically forbade agents from disclosing the comparative costs of the various modes of payment, though its Compliance Manual places strict requirements for pre-approval of any “sales material” proposed to be used by selling agents.
{7} At the hearing, the district court heard testimony from Plaintiff, her insurance broker, an expert on insurance regulation as it affects consumers who testified for Plaintiff, a marketing consultant employed by Old Line, an expert in life insurance marketing who testified for Old Line, and an expert in insurance regulation who testified for Old Line.
{8} The facts specific to Plaintiffs case are straightforward. She contacted an independent insurance broker inquiring about term life, disability, and medical insurance. The broker secured a policy for Plaintiff with Old Line. Plaintiff asserts that the broker never told her there was a different premium depending on the mode or frequency of payment she chose. Plaintiff also admitted that, until just prior to filing this suit, she did not do the arithmetic to determine if her premium she was paying exceeded the “maximum” premium reflected on the face of the policy.
{9} Plaintiffs insurance broker testified generally that it was his practice to explain the dollar difference between an annual premium and the mode chosen by his clients. He could not recall specifically if he had that conversation with Plaintiff, or whether he reviewed the actual policy with her after it was issued. He agreed he did not give comparisons for all payments and that he did not disclose modal factors as such. The sales brochure describing premium calculations was not to be distributed to the public. The insurance agent agreed he never gave Plaintiff the modal factors for her policy, and in fact he didn’t know the modal factors and didn’t care about them. His focus was on periodic cost. The agent agreed he did not disclose modal premiums in terms of an APR or interest rate. The agent said “I’d confuse her and myself.”
{10} Old Line’s marketing consultant— who has never actually sold life insurance— testified generally that Old Line does not provide a script to be used during sales and it does not try to teach agencies “how to actually sell life insurance.” She also testified that agents are not required to get prior approval of oral statements to prospective policyholders. The marketing consultant agreed that Old Line’s training materials are standard across the country and include rate quotation software that does not disclose dollar differential between modes and does not *401give modal factors. The marketing consultant also agreed that there was nothing in Old Line’s training materials telling agents they should disclose modal factors, dollar differences between payment modes, or interest rate calculations.
{11} Old Line’s marketing and sales expert opined that Old Line’s life insurance products are sold in unique, individual transactions. He likened life insurance sales to a financial transaction where people’s “wants and needs and affordability and sophistication” differ. In his opinion it would be necessary to “interview” both the agent and the buyer to find out what happened in each transaction. He saw nothing in the standard agent’s agreement that limited what an agent could tell a policyholder concerning premiums. This expert gave no opinion concerning the relative utility of disclosing dollar differentials between modes, or modal factors, or interest rates and APR equivalents.
{12} Twelve days after the close of the hearing, the district judge convened the parties in order to issue her oral ruling.1 The district court divided her remarks fairly neatly, dealing separately with the issues raised by the nationwide nature of the class, the issues surrounding factual predominance, and the Rule 1-023(A) NMRA issues.2
{13} The district court conducted an evaluation of the state of the law across the country in accordance with the requirements of Sun Oil Co. v. Wortman, 486 U.S. 717, 730-31, 108 S.Ct. 2117, 100 L.Ed.2d 743 (1988) and Phillips Petroleum Co. v. Shutts, 472 U.S. 797, 815, 105 S.Ct. 2965, 86 L.Ed.2d 628 (1985). The district court concluded, after reviewing the loss surveys provided by the parties, that the law was sufficiently uniform to allow class treatment of the breach of contract, duty of good faith, and the duty to disclose material facts causes of action. The district court recognized there were some variations among the states with regard to the duty of good faith claim, but proposed to address them by creating subclasses. It is unclear how the district court proposed to structure the subclasses or how they would actually be handled at trial. The district court also perceived some variation in breach of contract law in a few states, but felt it would not defeat predominance because the differences — involving the reasonable expectations doctrine — were not material to this ease.
{14} With regard to factual predominance, the district court decided that common issues regarding Old Line’s standardized forms training procedure and conduct would prevail over any potential individual issues. The district court recognized there were disputed issues of fact with regard to the disclosures made to Plaintiff. However, the district court felt that the case was unlikely to devolve into an exploration of each policy sales transaction. The district court in fact predicted that extrinsic evidence would not be allowed into evidence on the breach of contract claim.
{15} The district court was much less sanguine about whether the duty to disclose material facts claim would evade individualized analysis of cases, but certified it on a “when in doubt — certify” approach.
ANALYSIS
{16} We apply an abuse of discretion standard of review to class certification orders. “If the district court has applied the correct law, we will uphold its decision if it is supported by substantial evidence.” Berry, 2004-NMCA-116, ¶ 25,136 N.M. 454, 99 P.3d 1166.
BREACH OF CONTRACT AND DUTY OF GOOD FAITH
{17} We will deal with the breach of contract and duty of good faith claims summarily because the legal and factual issues here are indistinguishable from those we just decided in Berry. In Berry, we provided a comprehensive analysis of New Mexico insurance breach of contract law and its relationship to other states’ law on the issue. See Berry, 2004-NMCA-116, ¶¶ 82-88, 136 N.M. 454, 99 P.3d 1166. That analysis, including potential parameters for litigation of the claim, applies with full force here and need *402not be repeated, in particular since the district court here has not dealt with any issues going to the merits of the case. We are unconcerned with the district court’s statement that extrinsic evidence would not be allowed. We view the statement as more of a prediction (which may or may not prove accurate) for the course of litigation than as a ruling on the law in the case. Thus, we conclude, as we did in Berry, that the law of breach of contract “is uniform enough that our traditional notions of fair play and justice would not be offended by litigating the issue under New Mexico law.” Berry, 2004-NMCA-116 ¶ 82.136 N.M. 454, 99 P.3d 1166. We therefore affirm the district court’s certification of this claim.
{18} Berry also provides a full analysis of the duty of good faith issue, which applies here. The one difference between the cases is that the district court here did recognize that subclasses would be necessary to take state law variations into account. However, the district court did not take the analysis further and determine how many subclasses might be necessary. As a result, the district court did not explicitly gauge the effect of the subclass strategy on the question of predominance and jury trial management. As we stated in Berry, the parties and the district courts should be able to construct a reasonably definite trial plan with regard to these issues when certification is decided. Id. ¶ 96. We therefore remand the certification of the duty-of-good faith claim for the district court’s consideration of these matters.
{19} In addition, we note that the list of states with “different” duty-of-good-faith law, which the district court below noted, is different from the list we compiled in Berry. Id. ¶¶ 90, 91, 92. We will not try to reconcile the differences here since we are remanding for further consideration. We do note that the list in Berry was not intended to be comprehensive. The specific discussion there was in response to defendant Kemper’s argument and briefing.
FAILURE TO DISCLOSE
{20} The failure to disclose count poses a new problem for us. The district court in Berry refused to certify a similar claim and the plaintiff there did not appeal the decision. Thus, we have not had occasion to compare New Mexico’s law in the area with such law across the country. As we made clear in Berry, New Mexico courts cannot apply our law to parties from other states unless we can reasonably assure ourselves that the law in the affected jurisdictions is sufficiently consonant to avoid constitutional problems. Id. ¶¶ 77-80.
{21} The district court reviewed the law surveys provided by the parties. The Judge’s reaction to the failure to disclose law was two-fold. First, she acknowledged that “12 states seem to say that there’s not a fiduciary relationship between the insured and the insurance company.” She deflected that concern by observing that “those states do not stand for the proposition that there’s no duty to disclose material facts by an insurance company to an insured regarding premiums.” Thus, she found no conflict. Second, the district court also agreed that “[t]here’s lots of conflict in the area of fraud.” But, in the district court’s view, Plaintiff did not plead fraud and thus even “material differences and conflicts between the laws of New Mexico and the laws of some of the other states with respect to fraud is immaterial.” The district court did not recognize a need for subclasses in connection with this claim.
{22} Old Line argues that the failure to disclose claim is inescapably a fraud claim and simply is not a candidate for any class treatment—much less a nationwide class— because of the basic proof requirements for the tort and because of differences in the law across the country. Alternatively, Old Line argues that in any event the failure to disclose claim requires finding a duty to disclose, and asserts that the law regarding such a duty is too variable to sustain a nationwide class.
{23} Plaintiffs response is three-fold. Perhaps sensing some weakness in her position, Plaintiff first observes that the failure to disclose claim is simply an alternative claim which may not have to be dealt with at all, at least not until the breach of contract claim is resolved. We reject this suggestion. All class action certification issues should be dealt with on their merits by the district and *403appellate courts when and as presented. We disavow any approach which allows certification decisions to be made on a simple “we can deal with it later” basis. We certainly will not postpone consideration of the issue and chance, if not invite, another interlocutory appeal later.
{24} Plaintiffs more substantive response is that there are no individualized fact issues undermining the district court’s finding of predominance, and that there are no material differences in the law of the various states. The district court did not specifically address the potential factual issues surrounding the failure to disclose claim. Because we can resolve the claim on other grounds, neither will we.
{25} We disagree that the law is sufficiently uniform to sustain a nationwide class action and reverse on that basis. In any consideration of a multistate class action, the district court’s first task should be to decide what New Mexico’s law is on the subject. Only then can comparison be profitably made to other states.
{26} In Azar v. Prudential Insurance Co., 2003-NMCA-062, 133 N.M. 669, 68 P.3d 909, another case involving modal premiums, we explored in detail the contours of an insurer’s “duty to disclose.” We made clear that in New Mexico insurers have no pre-issuance duty to disclose based on the covenant of good faith or any notion of fiduciary duty to the insured. Id. ¶¶ 53, 56. We held that a pre-issuance common law duty might exist based on a series of cases recognizing an obligation on insurers and insureds “not to misrepresent or withhold information material to an insurance contract.” Id. ¶¶ 57, 62, citing Modisette v. Found. Reserve Ins. Co., 77 N.M. 661, 667, 427 P.2d 21, 25 (1967). The “general rule” recognized by the court in Modisette was not dependent on a finding of fraudulent intent or even negligent conduct for application. The duty is apparently grounded in the materiality of the information misrepresented or omitted. “To be material, the false statement does not have to actually contribute to a loss under the terms of the policy.” Prudential Ins. Co. v. Anaya, 78 N.M. 101, 104, 428 P.2d 640, 643 (1967). Materiality can be presumed in appropriate circumstances. See Rael v. Am. Estate Life Ins. Co., 79 N.M. 379, 382, 444 P.2d 290, 293 (1968).
{27} Thus, in New Mexico an insured does not have to rely on fraud to state a claim for a failure to disclose material information, though she can assume the burden of relying on fraud if she chooses. Plaintiff converts this state of the law in New Mexico into a general duty to disclose and argues by extension that for purposes of certification the contours or sources of this duty do not matter; what is important is the mere existence of the duty.
{28} We disagree. How a state imposes its duty to disclose — its doctrinal base and evidentiary requirements — cannot simply be ignored. If another state requires a showing of classic fraud to make a claim for failure to disclose, it would be constitutionally improper for New Mexico to impose a duty and consequent damages on a lesser showing. Wortman allows New Mexico to apply its law to transactions occurring in other states only if our law is substantially similar to the other jurisdictions. Thus, the district court was wrong when it decided that the contours of the law creating a duty to disclose were irrelevant. This was a basic error of law which under our standard of review supports, if not requires, a finding of abuse of discretion compelling reversal.
{29} While we certainly do not prejudge the issue, we find it difficult to see how the district court will be able to construct an acceptable class for this cause of action. We are convinced that a nationwide class decided under our Modisette fine of cases is impossible. Plaintiff has not even attempted to demonstrate that the Modisette rule is recognized across the country. For all we know it is a small minority rule. Lacking a basis for a nationwide class, the district court would have to try and construct a manageable set of subclasses. A subclass limited to Modisette-type jurisdiction might be workable. We do not believe, however, that the district court could properly try to create subclasses of states that require any type of fraud as the basis for the duty to disclose. As Plaintiff reminds us, she has not attempted to state a claim for fraud as the basis for her duty to disclose a cause of action. We do not *404see how Plaintiff could properly be a representative for and conduct a class in which fraud would have to be shown as a predicate to finding the duty.
CONCLUSION
{30} We affirm the certification of a class action with regard to the breach of contract claim. We reverse the certification of the duty of good faith and duty to disclose claims. The case is remanded for further proceedings not inconsistent herewith.
{31} IT IS SO ORDERED.
WE CONCUR: A. JOSEPH ALARID, Judge and CYNTHIA A. FRY, Judge.