Vos v. Farm Bureau Life Insurance Co.

667 N.W.2d 36, 2003 Iowa Sup. LEXIS 143, 2003 WL 21659083

• Court: Supreme Court of Iowa
• Decided: July 16, 2003
• Docket: 01-1025
• Status: Published

Opinion

LAVORATO, Chief Justice.

The plaintiffs, Cherri Lea Vos, Rita C. Anderson, and Quentin V. Anderson, appeal a district court ruling decertifying their class action against Farm Bureau Life Insurance Company and FBL Financial Group, Inc. (collectively “Farm Bureau”). Because we find no abuse of discretion in the court’s ruling, we affirm.

I. Background.

A. The allegations. On February 1, 1999, Vos and her parents, the Andersons, filed a class action against Farm Bureau. The plaintiffs alleged that from January 1, 1982 through at least December 31, 1996, Farm Bureau engaged in a pattern of deceptive practices in connection with the sale of whole life and universal life polices. The deceptive conduct as alleged in the petition involved:

(a) Selling whole and universal life insurance policies under the pretense of “vanishing premiums” and/or “fully paid up” policies. Farm Bureau promised its existing policyowners and new customers that, on a given policy, they would not have to make additional premium payments after a specified number or amount of payments or years (“vanishing premiums”); and

(b) Strongly urging policyowners to surrender, borrow, or withdraw cash values from their existing policies to fund an additional policy, or an entirely new Farm Bureau policy which replaced the existing policy, usually with the added incentive that premiums on the second “fully paid up” policy would be paid substantially or entirely from the value of the first policy and at no additional cost to the policyowner (“replacement of policies with vanishing premiums”).

The petition further alleged the following additional facts. On May 17, 1988, Farm Bureau agent Merrill Perry met with plaintiffs at their home to “induce them into purchasing an additional ‘fully paid up’ life insurance policy” for Cherri Vos. Perry used policy illustrations and calculations provided by Farm Bureau to represent to the plaintiffs this “great deal” would allow them to obtain a fully paid up $50,000 special whole life policy on Cherri without making further premium payments on the policy. Perry convinced Rita to cash in a $2,000 paid up life insurance policy she owned through another insurance company and apply the proceeds to the policy for Cherri. In addition, Perry persuaded Quentin to write a check in payment of the additional amount necessary to make the new policy “fully paid up.” In December 1998, Farm Bureau offered to recognize Cherri’s policy as “fully paid up,” provided she agree to accept a reduced paid up policy in the amount of $20,710.50.

The petition alleged the following common questions of law and fact predominated over individual questions:

(1) Whether Farm Bureau, through its national general agency system and sales force of agents, engaged in deceptive acts and practices in connection with the sale of whole life and/or universal life insurance policies; and

(2) Whether Farm Bureau, through its national general agency system and sales force of agents, breached its contracts with plaintiffs and the members of the class during the class period by requiring additional out-of-pocket payments on whole and/or universal life insurance policies purchased by plaintiffs and members of the class beyond the amounts or dates represented.

The petition asserted five separate claims against the defendants: breach of contract, negligent misrepresentation, common law fraud, fraudulent inducement, and breach of fiduciary duty. The defendants answered and asserted several affirmative defenses. One of those affirmative defenses is relevant to this appeal: There are no common questions of law or fact which predominate over any questions affecting only individual members.

B. Proceedings related to class action certification. On July 15, 1999, the plaintiffs filed a motion for class certification and a memorandum of law in support of that motion, which Farm Bureau resisted. In its memorandum of law the plaintiffs asserted that the defendants engaged in a uniform and common course of unfair and deceptive conduct that was substantially similar with respect to each member of the proposed class. According to the plaintiffs,

[T]he deceptive acts and practices challenged by plaintiffs are based on uniform sales practices whereby Farm Bureau failed to inform existing policy-owners and new customers of the substantial risks associated with these insurance policies, i.e., that the only way in which premiums would “vanish,” or would not have to be paid on the additional or new policies, was if interest rates and/or dividend scales remained relatively constant. Defendants also failed to inform their policyowners and new customers, inter alia, that if the interest rates and/or dividend scales declined, by even a fraction of a percent from that used as the basis for the agents’ representations and inducements, then the specified number or amount of payments and/or the dividends and/or interest on the policies would be wholly insufficient to pay for the additional premiums and/or “fully pay up” the policies, and policyowners would have to pay the difference out-of-pocket.

Following a hearing in which the district court heard arguments on the motion, the court granted the plaintiffs’ motion for class certification on February 14, 2000. In ruling on the motion, the court considered the petition and some thirty exhibits attached to the plaintiffs’ memorandum. In its ruling, the court noted that “the class certification order is always subject to modification should later developments during the course of the trial so require.” The district court defined the class as follows:

All persons who have or had an ownership interest in one or more whole life and/or universal life policies issued during the period of January 1, 1982 through December 31, 1996 (the “class period”) that was sold by Farm Bureau through its exclusive agency force, and to whom it was represented that premiums would “vanish” and/or that a policy would be “fully paid up” and/or that no additional costs would be incurred in excess of the specified number or amount of payments or years. Excluded from the class are defendants, any parent, subsidiary or affiliate of defendants, any entity in which defendants have a controlling interest, and the officers, directors, agents, employees, legal representatives, heirs, predecessors, successors, and assigns of such excluded person or entity.

C. Discovery. Discovery commenced in June 1999 and continued for nearly two years during which Farm Bureau produced over 60,000 pages of documents and the plaintiffs produced over 100 documents. The plaintiffs as well as several of Farm Bureau’s present and former employees and senior management officials were deposed. The documentation Farm Bureau produced included sample life insurance contracts, sample illustrations used by Farm Bureau agents, minutes of board of directors’ meetings concerning interest rates and dividend declarations, brochures and promotional materials, training manuals, rate manuals, complaint files, Farm Bureau’s SEC Prospectus regarding its public offering, actuarial material regarding the policies in question, organizational charts, and affidavits. The documentation the plaintiffs produced included the plaintiffs’ life insurance policies and attendant documents, copies of checks regarding premium payments, and documents regarding the plaintiffs’ consumer complaint.

The discovery reveals the following information about the type of policy at issue and the marketing scheme Farm Bureau used to sell it.

1. The policies. The plaintiffs purchased identical policies from Farm Bureau: “Special Whole Life Participating.” Quentin and Rita Anderson purchased their policies in 1985. The allegations of the petition are limited to the facts surrounding the purchase of Cherri’s policy, which was purchased in 1988. Therefore the facts surrounding the purchase of Rita’s and Quentin’s policy are relevant only to the extent they are included as class members.

The front page of the policy contained the following:

SPECIAL WHOLE LIFE POLICY

PARTICIPATING

Payable at Death.

Level Premiums Payable for Life.

Annual Dividends.

[[Image here]]

RIGHT TO CANCEL

The owner may cancel this policy by ... returning the policy or contract before midnight of the twentieth day after the date you receive the policy.... Farm Bureau Life will refund any premium paid for this policy within ten days after it receives notice of cancellation and the returned policy.

(Emphasis added.)

The top of the table-of-contents page informed the policyholder, “This policy is a legal contract between the policy owner and the Farm Bureau Life Insurance Company. READ YOUR POLICY CAREFULLY.”

The “Policy Data Sheet” is found at page three of the policy. It lists the “annual premium” and states that the “period payable” is “lifetime.” The policy data sheet also lists the guaranteed cash value and paid-up insurance the policy will generate or carry at any given point in time.

The policy contained the following provisions related to premiums and dividends, and the following general provisions:

PREMIUM PROVISIONS

Premium Payment. Premiums are to be paid for the term shown on page 8 [Policy Data Sheet], The amount and frequency of premium payments are also shown on page 3.

DIVIDENDS PROVISIONS

Dividend Participation. As long as this policy is in force, except as extended term insurance, it will share in the company’s surplus as set by the company each year. The share will be paid as a dividend at the end of each policy year. However, no dividend will be paid until the second policy year.

Dividend Options. The owner has the following dividend options:

1) Cash- — Dividends will be paid in cash.

2) Premium Reduction- — Dividends will be used to help pay premiums. Any balance will be paid in cash.

3) Accumulations- — Dividends will be left with us to earn interest. We will set the interest rate, but it will not be less than 3% per year.

4) Paid-Up Additions — Dividends will be applied to buy paid-up additional insurance. The insurance will be of the same type as this policy. Paid-up additions will also share in the company’s surplus.

The owner may choose a dividend option on the application of this policy or by a later written request to us. If a dividend option has not been chosen, dividends will be applied under the paid-up additions option.

GENERAL PROVISIONS

Contract. This policy is issued in consideration of the first premium and the statements in the application. This policy, and the attached application and all other attached papers, if any, which by this reference are made a part of this policy, form the entire contract. All statements in the application are deemed representations and not warranties except in the case of fraud. No statement will void this policy or be used in defense of a claim unless contained in the application.

Modification. No one can change any part of this policy except the owner and one of our officers. Both must agree to a change, and it must be in writing. No agent may change this policy or waive any of its provisions.

Farm Bureau issued a “Premium Deposit Agreement” to each plaintiff in 1988. Such an agreement is issued to each policyholder upon receipt of a sum of money that represents a number of annual premiums on a life insurance policy. The agreements the plaintiffs received acknowledge the receipt by Farm Bureau of a sum of money representing a number of annual premiums on their policies. The agreements explain, “This deposit will be used in prepayment on the above policy of the annual premiums specified above.” The agreements indicate four annual premiums prepaid for Quentin, three for Rita, and nine for Cherri.

2. Marketing the policies. Farm Bureau markets its life insurance products through an exclusive agency force. Its agents are independent contractors and exclusive agents of Farm Bureau. This is in contrast to other insurance agents, who may sell insurance products of several different companies.

Farm Bureau provided its agents with software to create illustrations during sales presentations to clients. Before 1988, Farm Bureau agents produced illustrations by inputing information into a hand-held computer. The computer generated a piece of paper resembling an adding machine tape. When interest rates changed, agents would receive a new chip to insert into the hand-held computer. The chip was programmed with the then-current declared interest rate to be used by Farm Bureau agents in calculating dividends.

After 1988, agents used laptop computers and received floppy disks when interest rates changed. An agent could not prepare an illustration using an interest rate higher than that which the board of directors had declared as that period’s rate for use in calculating dividends. Agents could not input an interest rate. The interest rate was hard-coded in the chip (and later the software).

Agents could present one of five illustration options for Special Whole Life insurance: (1) dividend to reduce premium, (2) dividends to buy paid-up additions, (3) dividend to accumulate with interest, (4) limited pay concept, and (5) rated policy. All illustrations contained the following language:

This illustration is based on the company’s current dividend scale which is sensitive to changing interest rates. Dividends are neither guaranteed nor estimated for the future.

Steven Knutzen, a Farm Bureau employee who trained Farm Bureau agents, explained in an affidavit that

the illustration system was capable of calculating how many advance premiums would need to be paid on a policy to get it to the point where, at the then-current interest rates, it would generate dividends at such a level that the dividends would pay premiums from that point into the future....

We explained to agents, and we expected and trained them to explain to clients, that just because no cash premium payments were being made [in the hypothetical being presented] that it did not mean premiums were not being paid. Premiums still came due, but they were paid out of the cash value of the paid-up additions (dividends). The client knew this every year, because every year the client would receive a form which he/she was required to sign before Farm Bureau was allowed to withdraw any monies from the dividend fund to pay the annual premium.

According to Knutzen, the Universal Life illustration system allowed an agent to run three separate illustrations using three different interest rates: the then-current rate (12%), the guaranteed rate (4.5%), and a rate midpoint between the guaranteed rate and the then-current rate (9%). Agents were trained to project all three rates, because everyone knew the then-current rate was not going to continue. Farm Bureau wanted their agents to be realistic with their clients in terms of how these policies could .reasonably be expected to perform. Agents were also trained to communicate to their clients that the illustrations were projections based on the current dividend scale, that dividends were not guaranteed and that the interest rate used to calculate them was not guaranteed. Agents were instructed to leave illustrations with prospective clients after discussing them. Farm Bureau, however, did not require its agents to use illustrations or other materials it provided for use in sales presentations.

3. Dividends/interest rate declarations. Farm Bureau declares dividends each year. Dividends are a function of three factors: a gain from interest, a gain from mortality, and a gain .from expenses. Dividends are a distribution of the gain that the company has experienced because actual performance turns out better than the pricing assumptions. Even if the rate does not change from one year to the next, the rate must be redeclared by the board of directors.

Farm Bureau’s policies state, “[a]s long as this policy is in force, ... it will share in the Company’s surplus as set by the company each year.” Special Whole Life policies do not earn dividends, if any, until the second full year of the policy.

Dividend interest rates declared by Farm Bureau’s board of directors rose from 9 percent to 10.5 percent in 1984, to 12 percent in 1985, then declined to 11 percent in 1986 and 9 percent in 1987. Rates remained at 9 percent until 1992. The first credited dividends for policies issued in 1981 or 1982 were at rates higher than rates illustrated by the company. In contrast, the first credited dividends for policies issued in 1985 and 1986 were significantly lower.

D. Proceedings related to class action decertification. On February 26, 2001, Farm Bureau filed a motion to decer-tify the class. The sole ground urged in the motion was that “the individual questions and issues will always predominate over the common.” Following a hearing on the motion, the district court granted the motion to decertify. In its ruling, the court concluded:

[F]or each cause of action forwarded by plaintiffs, ... individual questions of law and fact predominate common questions .... [A]fter nearly a full year of discovery, plaintiffs cannot direct this court to any piece of evidence to support their theory of uniform deception.... [T]he individual facts behind each claim [are] more case specific than common.

The plaintiffs appealed from this ruling.

We set out additional information as that information relates to the issues raised in this appeal.

II. Issues.

The plaintiffs contend the district court abused its discretion in decertifying the class action in light of substantial evidence that common issues of law and fact predominate. Whether the district court abused its discretion as the plaintiffs contend turns on our determination whether common questions of law or fact predominate over any questions affecting only individual class members.

III. Scope of Review. Our review of the district court’s ruling granting or denying certification of a class action is limited because the district court enjoys broad discretion in the certification of class action lawsuits. Stone v. Pirelli Armstrong Tire Corp., 497 N.W.2d 843, 845 (Iowa 1993). We will reverse an order granting or denying certification only if we find the district court’s decision was based upon an abuse of discretion. Varner v. Schwan’s Sales Enters., Inc., 433 N.W.2d 304, 305 (Iowa 1988). We find such an abuse of discretion only where the district court’s grounds were clearly unreasonable. Id.

The same broad discretion exists in determining whether to decertify a class. 5 James Wm. Moore et al., Moore’s Federal Practice § 23.04, at 23-28 to 23-29 (3d ed.2002) (citing In re Int’l House of Pancakes Franchise Litig., 536 F.2d 261, 263 (8th Cir.1976)). See also Iowa R. Civ P. 1.265(1) (providing that in amending the certification order, the court may (1) establish subclasses, (2) eliminate from the class any class member who was included in the class as certified, (3) provide for an adjudication limited to certain claims or issues, (4) change the relief sought, or (5) make any other appropriate change in the order).

IV. Applicable Law.

Iowa Rules of Civil Procedure 1.261 to 1.263, the rules regarding class actions, closely resemble Federal Rule of Civil Procedure 23. Compare, e.g., Fed.R.Civ.P. 23(b)(3), with Iowa R. Civ. P. 1.263(e), (⅞)-(⅞). Therefore, we may rely on federal authorities construing similar provisions of Federal Rule of Civil Procedure 23. Vignaroli v. Blue Cross of Iowa, 360 N.W.2d 741, 743 (Iowa 1985).

Iowa Rule of Civil Procedure 1.261 permits the commencement of a class action if there is a question of law or fact common to a class of persons so numerous that joinder of all persons is impracticable. Iowa R. Civ. P. 1.261. The district court may certify an action as a class action if it finds:

a. The requirements of rule 1.261 [nu-merosity and commonality] have been satisfied.

b. A class action should be permitted for the fair and efficient adjudication of the controversy.

c. The representative parties fairly and adequately will protect the interests of the class.

Iowa R. Civ. P. 1.262(2).

The plaintiff has the burden of establishing that a purported class of plaintiffs meets the prerequisites. Vignaroli, 360 N.W.2d at 744. A failure of proof on any one of the prerequisites is fatal to class certification. City of Dubuque v. Iowa Trust, 519 N.W.2d 786, 791 (Iowa 1994).

Rule 1.263(1) sets forth a number of criteria the district court should consider in determining whether the prerequisite of rule 1.262(2)(5) (the class action should be permitted for the fair and efficient adjudication of the controversy) has been met. The criteria center on two broad considerations: “achieving judicial economy by encouraging class litigation while preserving, as much as possible, the rights of litigants' — -both those presently in court and those who are only potential litigants.” Vignaroli, 360 N.W.2d at 744. The rule does not require the district court to assign weight to any of the criteria listed, “further evidencing an intent to grant considerable discretion to the trial court.” Id. Nor does the rule require the court to make written findings as to each factor under rule 1.263(1). City of Dubuque, 519 N.W.2d at 793. Rather the rule merely requires the court to weigh and consider the factors and come to a reasoned conclusion as to whether a class action should be permitted for a fair adjudication of the controversy. Id.

One of the criteria in rule 1.263(1) is that “common questions of law or fact predominate over any questions affecting only individual members.” Iowa R. Civ. P. 1.263(l)(e). It is this criterion that is primarily at issue in this case. Inherent in the court’s inquiry into this issue

is the recognition [that] the class action device is appropriate only where class members have common complaints that can be presented by designated representatives in the unified proceeding. The question as to whether common issues of fact or law predominate over those affecting only individuals is a fairly complex one.

Vignaroli, 360 N.W.2d at 744.

The individual claims need not be carbon copies of each other. Id. at 745. If a “common nucleus of operative facts” is present, a class action can be brought even without a “complete identity of facts relating to all class members.” Id. (quoting 7A C. Wright & A. Miller, Federal Practice and Procedure § 1778, at 53-54 (1972)). “ ‘Predominate’ should not be automatically equated with ‘determinative’ or ‘significant.’ ” Id. “A claim will meet the predominance requirement when there exists generalized evidence which proves or disproves an element on a simultaneous, class-wide basis, since such proof obviates the need to examine each class member’s individual position.” Cope v. Metro. Life Ins. Co., 82 Ohio St.3d 426, 696 N.E.2d 1001, 1004 (1998) (citation omitted).

Certification of a class action does not depend on a determination of whether the plaintiffs will ultimately prevail on the merits. Vignaroli, 360 N.W.2d at 745. Rather, it depends on whether or not the requirements of the rule governing class actions are met. Id. “The appropriate inquiry is not the strength of each class member’s personal claim, but rather, whether they, as a class, have common complaints.” Martin v. Amana Refrigeration, Inc., 435 N.W.2d 364, 367 (Iowa 1989). For this reason, the court does not conduct a preliminary inquiry into the merits of a suit in a certification hearing. Id. However, “that is not to say that the court may not require sufficient information to form a reasonable judgment in deciding whether to certify a class action.” Id. at 367-68; see also Cohn v. Mass. Mut. Life Ins. Co., 189 F.R.D. 209, 212 (D.Conn. 1999) (“Because ‘the class determination generally involves considerations that are enmeshed in the factual and legal issues comprising the plaintiffs cause of action,’ the Rule 23 analysis often requires the court ‘to probe behind the pleadings before coming to rest on the certification question.’ ” (Citations omitted.)); In re Jackson Nat’l Life Ins. Co. Premium Litig., 183 F.R.D. 217, 219 (W.D.Mich.1998) (“Maintainability may be determined by the court on the basis of the pleadings, if sufficient facts are set forth, but ordinarily the determination should be predicated on more information than the pleadings will provide.” (Citation omitted.)); Kramersmeier v. R.G. Dickinson & Co., 440 N.W.2d 873, 877 (Iowa 1989) (“[A] preliminary inquiry into the legal and factual sufficiency of the plaintiffs’ claim is appropriate inasmuch as they seek to represent the interests of the entire class.”)

In particular, the question of predominance necessitates a “close look” at “the difficulties likely to be encountered in the management of a class action.” Rothwell v. Chubb Life Ins. Co. of Am., 191 F.R.D. 25, 28-29 (D.N.H.1998) (quoting Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 614-17, 117 S.Ct. 2231, 2246, 138 L.Ed.2d 689, 707-09 (1997)). Moreover, “[t]he duty of the trial court to ensure compliance with Rule 23(a) continues after certification of the class, and the court may decertify the class after the initial certification if appropriate.” In re Hartford Sales Practices Litig., 192 F.R.D. 592, 602 (D.Minn.1999). See also Roby v. St. Louis Southwestern Ry., 775 F.2d 959, 963 (8th Cir.1985) (holding that district court did not abuse its discretion in decertifying class following trial on the merits).

Y. Class Certification of “Vanishing Premium” Cases, Generally.

Federal courts have addressed the issue of class certification in “vanishing premium” .cases over the past several years. An examination of these cases and the general trend in the federal (and, to a certain extent, state) courts give us a general framework from which to begin our analysis of the issue before us.

A “vanishing premium” policy

is one where, after a certain number of premium payments, the policy itself generates sufficient income through dividends and interest to pay any additional premiums due. In other words, premiums paid in the initial years of a policy are supposed to generate sufficient value to pay all future premiums due in later years, so that future premiums, in essence, “vanish.” The assumption that premiums will “vanish” depends on, among other factors, mortality experience, expenses, dividends, and associated interest rates. A change in any of these variables can affect the date by which premiums will “vanish.”

Parkhill v. Minn. Mut. Life Ins. Co., 188 F.R.D. 332, 335 n. 2 (D.Minn.1999). These types of policies became popular in the early 1980s, when interest rates hit record highs. See generally Daniel R. Fischel and Robert S. Stillman, The Law and Economics of Vanishing Premium Life Insur ance, 22 Del. J. Corp. L. 1, 3-8 (1997) (explaining the economics of the vanishing premium problem).

As one writer indicates, plaintiffs have generally been unsuccessful in certification motions involving vanishing premium cases based primarily on the issue of predominance of common issues to individual ones. See generally James R. Carroll, Recent Developments in “Vanishing Premium,” SF81 ALI-ABA 1 (2001) (collecting cases) [hereinafter “Carroll”]. Carroll cites three cases that “have established a body of precedent that is almost always followed in the adjudication of motions to certify class actions.” Id. at 2. They include Cohn, 189 F.R.D. at 211, 218 (class certification denied as to claims for breach of contract, negligence, fraud, negligent misrepresentation, unjust enrichment and violations of the Connecticut unfair trade practices statute); Parkhill, 188 F.R.D. at 340-45 (class certification denied as to claims for breach of contract and violations of Minnesota consumer protection statutes); and Jackson National, 183 F.R.D. at 219, 220-23 (class action certification denied as to claims for fraud, negligent supervision, breach of contract, unjust enrichment and statutory violations). In each of these cases the court found that the claims all required case-specific individualized showings of the alleged misrepresentations of the agent and the alleged reasonable reliance of the policyholder. See Cohn, 189 F.R.D. at 218; Parkhill, 188 F.R.D. at 342-45; Jackson Nat’l, 183 F.R.D. at 220-23.

In Cohn, the plaintiffs asserted the predominance requirement was met because

the written illustrations produced by Connecticut Mutual and distributed to the company’s network of agents, or printed by the agents themselves using software provided by the company, were identical in all material respects in each of the transactions at issue.... [A]s a result of Connecticut Mutual’s centralized training program, the sales presentations used by the sales people in each transaction were uniformly misleading. The evidence concerning these allegedly uniform illustrations and sales presentations, the underlying training programs, and the state of mind of Connecticut Mutual in orchestrating the alleged scheme, and the legal determinations that will be based on this body of evidence, ... constitute the core of the controversy regarding each class member.

Cohn, 189 F.R.D. at 212.

Rejecting the contention that the predominance requirement was met, the court concluded:

[T]he predominance requirement is not satisfied. Though there are issues common to the members of the proposed class, including in particular the actions and state of mind of Connecticut Mutual in pursuing the vanishing premium marketing strategy, these common questions are overshadowed by individualized issues such as the nature of the oral representations or disclosures made by the agent or broker at the point of sale, the nature of any questions asked by the consumer, the content of any written illustrations or disclosures given to the consumer, the degree of care exercised by the consumer in reviewing any written illustrations and the policy instrument, the portions of the offer that were attractive to the consumer, the degree of the consumer’s financial sophistication and his or her understanding of the product. These individualized issues, which are essential to the determination of the claims of each class member, predominate over questions common to the class.

Id. at 218.

In Jackson National, ■ the court noted the following relevant facts:

Jackson National did not generally communicate directly with prospective consumers or policyholders. Communications were made primarily by independent insurance brokers; brokers who were not subject to and did not follow uniform policies regarding distribution of policy illustrations....

Neither were brokers required to follow uniform sales scripts.... [T]his freedom led to great variance in representations made by brokers; some explaining away and others even exacerbating any misleading tendencies the policy illustrations may have had.

Jackson Nat’l, 183 F.R.D. at 221. The court concluded that a “determination of whether and which illustrations were given to class members, and of the nature of oral representations made to them at the point of sale, ... are matters requiring individualized fact development.” Id. In addition, “the materiality of the allegedly misleading illustrations and plaintiffs’ reliance on them” were “fact issues requiring individualized treatment.” Id.

Other federal courts have reached similar conclusions. See, e.g., In re LifeUSA Holding Inc., 242 F.3d 136, 145-48 (3d Cir.2001) (class certified by district court based on pre-sale actions of LifeUSA was an abuse of discretion because commonality, predominance, and superiority requirements were not met); Keyes v. Guardian Life Ins. of Am., 194 F.R.D. 253, 257 (S.D.Miss.2000) (class certification of action alleging fraud in the sale of “vanishing premium” policies denied because facts did not support finding “that defendant’s policies were sold in a sufficiently uniform manner so as to justify a finding that common issues predominate”); Rothwell, 191 F.R.D. at 30-31 (motion to certify “vanishing premium” subclass denied where “even if plaintiffs were able to prove at trial that Chubb trained its agents to use the policy illustrations in a misleading manner, it still would not eliminate the need for a ‘mini-trial’ on each class member’s claim to determine the nature of the representations that were made in that case” and resolution of fraud-based claims would require proof “both that Chubb’s agents made misrepresentations and that the individual class members reasonably relied on those representations in purchasing their insurance policies”).

Against this backdrop, we proceed to our analysis and our determination of the merits.

VI. Analysis.

We first address the plaintiffs’ contention that the district court “imper-missibly made factual determinations concerning the conflicting evidence presented by plaintiffs and defendants in connection with the decertification proceedings.” We note the district court conceded that it could not inquire into the merits of the case in determining whether to grant certification. However, we agree with the district court this principle is not absolute. In reaching a decision on a motion for class certification, the court is not limited to a review of the pleadings. To the contrary, where the issues are not

plain enough from the pleadings to determine whether the interests of the absent parties are fairly encompassed within the named plaintiffs claim ... it may be necessary for the court to probe behind the pleadings before coming to rest on the certification question. Even after a certification order is entered, the judge remains free to modify it in light of subsequent developments in the litigation.

Gen. Tel. Co. of S.W. v. Falcon, 457 U.S. 147, 160, 102 S.Ct. 2364, 2372, 72 L.Ed.2d 740, 752 (1982) (emphasis added).

Given the vast amount of documents produced following the certification order and the depositions taken, we think it was proper for the district court to probe behind the pleadings on the issue of predominance. We conclude therefore that rather than making impermissible findings on the merits the district court simply examined the vast amount of evidence that was available to determine this issue.

Certain evidence surrounding plaintiffs’ purchase of insurance from Farm Bureau supports the district court’s conclusion that plaintiffs’ allegations remained merely speculative after many months of discovery. For example, the district court pointed to the fact that the plaintiffs could not recall their agent ever showing them an illustration in connection with the life insurance sale. As the district court explained, this fact undermines the plaintiffs’ allegation that “the inflated and unrealistic performance models (based on ‘teaser’ interest rates) employed within the policy illustrations reveal company-wide deception.” Moreover, the court noted that the plaintiffs produced no evidence that the agents uniformly employed said illustrations. As the district court noted, without proof of uniform deception, individual questions of fact and law predominate the common. Additionally, the record shows that Rita cashed in her $2,000 life insurance policy four years before she claimed Merrill Perry had persuaded her to do so.

A. Breach of contract. As to this claim, the district court concluded individual questions of fact and law predominate the questions common to the class:

As to plaintiffs’ claim for breach of contract (Count II), because there is no evidence of company wide deception or that the agents uniformly employed the illustrations in the sales calls, the issue of whether the illustrations became part of the insurance contract becomes a question individualized for each putative plaintiff.

The plaintiffs contend that this ruling is incorrect because the plaintiffs’ breach of contract claim is predicated neither on the existence of “company wide deception” nor the use of “illustrations in sales calls [or] whether the illustrations become part of the insurance contract.” Rather, the plaintiffs contend, their breach of contract claim is based solely on the issue of whether the defendants knowingly created the “reasonable expectation” in connection with the policies purchased by the plaintiffs and the class that premiums would “vanish” and/or the policy would be “fully paid up” and/or that no additional costs would be incurred in excess of the specified number or amount of payments or years. The plaintiffs complain that although they discussed their reasonable expectations theory in their resistance to the defendants’ motion to decertify the class, the district court failed to discuss it.

Farm Bureau counters by noting that from the beginning the plaintiffs have focused on the common course of deception pursued by Farm Bureau against all policyholders rather than on the expectations of the policyholders themselves. The reasonable expectations theory, Farm Bureau asserts, was first raised in the plaintiffs’ resistance to the motion to decertify. (Actually the record shows the plaintiffs first raised the issue in its reply to Farm Bureau’s resistance to the plaintiffs’ motion to certify.)

According to Farm Bureau, the shift in theory resulted because the plaintiffs were forced to create some sort of argument in support of their breach of contract claims, particularly after the district court expressed at the first hearing reservations about whether claims which incorporated the illustrations — all of which bore written disclaimers — could really support a finding against Farm Bureau.

Additionally, Farm Bureau contends the doctrine of reasonable expectations simply does not apply in this case.

We too have serious doubt that the doctrine of reasonable expectations applies in these circumstances. We have applied the doctrine to insurance coverage cases

to avoid the frustration of an insured’s expectations notwithstanding policy language that appears to negate coverage. It is a narrow doctrine that is primarily employed when the insurance coverage provided eviscerates terms explicitly agreed to or is manifestly inconsistent with the purpose of the transaction for which the insurance was purchased.

Monroe County v. Int’l Ins. Co., 609 N.W.2d 522, 526 (Iowa 2000) (emphasis added).

The doctrine is inapplicable if (1) an ordinary layperson would not misunderstand the policy’s coverage as to this occurrence, and (2) there were no other circumstances attributable to the insurer at the time the policy was negotiated and issued that would foster coverage expectations. Such circumstances may be shown if the policy exclusion: (a) is bizarre or oppressive, (b) eviscerates terms explicitly agreed to, or (c) eliminates the dominant purpose of the transaction.

Zaragoza v. West Bend Mut. Ins. Co., 549 N.W.2d 510, 515 (Iowa 1996) (citations omitted) (emphasis added). The doctrine applies only to representations made by the insurer at the time of policy negotiation and issuance. Id. Additionally, this requirement emphasizes the reliance interest upon which the doctrine is founded. Id. at 516.

Here the issue is not whether insurance coverage exists under the policy. Rather, the issue the plaintiffs are raising relates to the premium and dividend provisions of the policy. We hesitate to expand the doctrine to this situation, where coverage under the policy provisions is not in issue.

We agree with Farm Bureau that even were we to apply the doctrine of reasonable expectations outside the realm of coverage cases, there was no evidence that that the wording of the policies was bizarre or oppressive, or that there was an agreement to a term which is eviscerated by the policies, or that the dominant purpose of the policies — to provide monetary relief to the beneficiary of the deceased insured — was somehow eliminated.

Moreover, even if the doctrine were to apply, the plaintiffs face an insurmountable reliance issue. The party asserting the doctrine of reasonable expectations must show not only the expectations, but also that they were relied upon by the insurance purchaser in deciding to buy the policy. Id. The reliance showing requires an inquiry into each individual class members’ experience in purchasing insurance, as we vrill discuss further in the context of the plaintiffs’ negligent misrepresentation, common law fraud, and fraudulent inducement claims. To determine the “reasonable expectations” of any given class member, the individual inquiries surrounding each class members’ purchase of insurance would predominate over any common issues.

For all of these reasons, we conclude the district court did not abuse its discretion in concluding that individual questions predominate common questions on this issue.

B. Breach of implied covenant of good faith and fair dealing. Related to their breach of contract claim, the plaintiffs contend Farm Bureau breached its implied covenant of good faith and fair dealing when it failed to allocate “hundreds of millions of dollars of surplus every year during the class period” to “the ambiguous contracts purchased by the plaintiffs and the class to meet their ‘limited pay/vanishing premiums’ expectations created by” Farm Bureau.

Farm Bureau correctly asserts the “good faith and fair dealing” theory emerged for the first time in the plaintiffs’ resistance to the motion to decertify. Farm Bureau also asserts the theory only applies in the context of a bad faith claim for an insurance company’s failure to indemnify and defend its insured.

The district court did not specifically address this claim in its decertification ruling. However, as with the reasonable expectations claim, we likewise conclude the implied covenant claim has no application in these circumstances.

“A covenant is implied in an insurance contract that neither party will do anything to injure the rights of the other in receiving the benefits of the agreement.” Kooyman v. Farm Bureau Mut. Ins. Co., 315 N.W.2d 30, 33 (Iowa 1982). We have permitted claims against insurance companies for breach of the implied covenant of good faith by recognizing the tort of bad faith in third-party and first-party situations. Johnson v. Farm Bureau Mut. Ins. Co., 533 N.W.2d 203, 207 (Iowa 1995).

The tort of bad faith arises in situations where the insurer has denied benefits or has refused to settle a third-party’s claim against the insured within the policy limits. See, e.g., Gibson v. ITT Hartford Ins. Co., 621 N.W.2d 388, 396 (Iowa 2001) (“To establish a first-party bad-faith claim, ‘a plaintiff must show the absence of a reasonable basis for denying benefits of the policy and defendant’s knowledge or reckless disregard of the lack of a reasonable basis for denying the claim.’” (Citations omitted.) (Emphasis added.)); North Iowa State Bank v. Allied Mut. Ins. Co., 471 N.W.2d 824, 828 (Iowa 1991) (noting that the court has recognized “a bad-faith cause of action in an insurer’s representation of an insured in a third-party liability claim,” and that “an insurer may be responsible for the tort of bad faith in first-party situations in which the dispute involves the insured’s right to recover under the policy”).

The plaintiffs have neither alleged in their petition nor in subsequent filings that they are relying on bad faith. Furthermore, the tort of bad faith simply has no application in the circumstances of this case.

C. Breach of fiduciary duty. The district court did rule on this issue:

Finally, in regard to plaintiffs’ claim for breach of fiduciary duty (Count V), plaintiffs have again failed to establish beyond mere speculation that there are common questions of law and fact for the entire class. First, the relationship between insurance sales agents and customer does not, by its very nature, appear to necessarily give rise to a fiduciary relationship. Second, because the relationship between insurance agents and their clients is not necessarily a fiduciary relationship, the determination of whether a fiduciary relationship exists must be based upon the particular facts and circumstances of each individual case. Thus, the individual questions predominate the common questions in regards to plaintiffs’ claim of breach of fiduciary relationship.

(Citations omitted.)

The plaintiffs contend this ruling is not supported by the record. They further contend a fiduciary relationship may exist on a class-wide basis where, as here, the company selling the insurance is not just an insurance company, but a genuinely unique organization akin to a family. The plaintiffs cite to Farm Bureau’s admission that it owed its customers reasonable duties of disclosure and care, and testimony of Farm Bureau’s executive vice president regarding the duties of loyalty, honesty, and integrity owed by Farm Bureau to its customers.

Farm Bureau, not surprisingly, contends the district court was correct in its ruling on this issue. For reasons that follow, we agree.

A fiduciary relationship exists between two persons “when one of them is under a duty to act for or to give advice for the benefit of another upon matters within the scope of the relation.” Kurth v. Van Horn, 380 N.W.2d 693, 695 (Iowa 1986) (quoting Restatement (Second) of Torts § 874 comment a, at 300 (1979)). Some relationships, such as those between an attorney and client, guardian and ward, principal and agent, and executor and heir, “necessarily give rise to a fiduciary relationship.” Id. at 696.

Indicia of a fiduciary relationship include

the acting of one person for another; the having and the exercising of influence over one person by another; the reposing of confidence by one person in another; the dominance of one person by another; the inequality of the parties; and the dependence of one person upon another.

Id. (citation omitted). “Because the circumstances giving rise to a fiduciary duty are so diverse, any such relationship must be evaluated on the facts and circumstances of each individual case.” Id.

In Rothwell, a vanishing premium case, the plaintiffs claimed the defendant insurance company and its agents breached their fiduciary duty “by encouraging plaintiffs to replace their preexisting policies when it was not beneficial to their interests.” 191 F.R.D. at 32. The court denied class certification, explaining the fiduciary duty claim as follows:

In order to succeed with [the breach of fiduciary duty] claim, each plaintiff would need to prove: (1) the existence of a fiduciary relationship; and (2) that the policy replacement was not beneficial to his or her interests. Both showings, however, necessitate individualized inquiries that render certification inappropriate.

Determining whether Chubb owed a fiduciary duty to each plaintiff would require a fact-intensive inquiry into the nature of the relationship between policyholder and agent.

Id. (citations omitted); see also Kaser v. Swann, 141 F.R.D. 337, 341 (M.D.Fla. 1991) (“To show the existence of a fiduciary relationship the members of the class would have to prove that an exchange of trust and confidence occurred between each plaintiff and the [defendant]. This would require testimony from every [plaintiff] and, as such makes this case unsuited for class certification.”)

Because the existence of a fiduciary relationship here depends on the facts and circumstances of each individual case, we conclude the district court did not abuse its discretion in concluding that individual questions predominate common questions on this issue.

D. Negligent misrepresentation, common law fraud, and fraudulent inducement. As to these claims, the district court concluded:

In regard to the plaintiffs’ claims for negligent misrepresentation, fraud, and fraudulent inducement (Counts I, III, IV), questions of justifiable reliance become unique to each putative plaintiff. The fact issue of what individual agents may have represented to putative plaintiffs predominate questions common to the class. In other words, the true focus appears to be placed upon the representations made by individual agents rather than the representations (e.g., through sales material and agent training) made by Farm Bureau as a whole. Further, without some evidence beyond mere speculation that Farm Bureau engaged in a uniform practice of misrepresentations or omissions, this Court does not believe the entire class is entitled to a presumption of reliance.

The plaintiffs challenge the district court’s conclusion that “questions of justifiable reliance become unique to each putative plaintiff’ and that “what individual agents may have represented to putative plaintiffs predominate questions common to the class.” In support of their challenge, the plaintiffs make the following argument. Farm Bureau’s agents were as much in the dark as the plaintiffs about the company’s approval of false and misleading “teaser rates” and “illustration rates” and the substantial probability that such rates might never be paid. Additionally, there is no evidence that Farm Bureau ever advised its agents of these facts. In short, if Farm Bureau failed to disclose these facts to its agents, then its agents could only fail to disclose these same facts in their sales presentations to the plaintiffs and the class. Therefore, individual presentations by agents to plaintiffs and the class are not relevant to the common issues here. Because material omissions are involved, the plaintiffs and the class are entitled to a presumption of reliance.

The plaintiffs’ common law fraud, negligent misrepresentation and fraudulent inducement claims have one element in common: justifiable reliance. See McGough v. Gabus, 526 N.W.2d 328, 331 (Iowa 1995) (common law fraud); Barske v. Rockwell Int’l Corp., 514 N.W.2d 917, 924 (Iowa 1994) (negligent misrepresentation); Lamasters v. Springer, 251 Iowa 69, 72, 76, 99 N.W.2d 300, 302, 304 (1959) (fraudulent inducement). The plaintiffs apparently concede this proposition but steadfastly stand on their presumption of reliance theory grounded on material omissions.

Several vanishing premium cases have addressed the presumption of reliance issue. Those that have refused to apply the presumption did so because the evidence failed to establish essentially uniform misrepresentations. See, e.g., Keyes, 194 F.R.D. at 255-57 (plaintiffs alleged that all class members were victims of standardized misrepresentations and omissions when defendant used its software to create the uniform illustrations; court held that presumption of reliance not appropriate where the individual plaintiffs did not receive the same mix of information — some agents used the illustrations and others did not and there was no standardized sales pitch); Cohn, 189 F.R.D. at 216-17 (plaintiffs asserted defendant’s training video discouraged agents from disclosing possibility that dividend rates could decline in the future; court rejected plaintiffs presumption of reliance argument because there was no evidence that all the agents saw the video and then made uniformly deceptive sales presentations); Jackson Nat’l, 183 F.R.D. at 221 (plaintiffs asserted vanishing premium illustrations “premised on unsupported and unsustainable interest crediting rate assumptions” were centrally prepared in the home office and distributed to brokers; court held presumption of reliance not appropriate where brokers were not subject to and did not follow uniform policies regarding distribution of policy illustrations and brokers were not required to follow uniform scripts; therefore information contained in the illustrations was conveyed to consumers, if at all, in the context of varying oral representations). See also Cope, 696 N.E.2d at 1004 (“[Although having some common core, a fraud case may be unsuited for treatment of a class action if there was material variation in the representations made or in the kinds or degrees of reliance by the person to whom they were addressed.” (quoting 1966 Advisory Committee Notes to Fed.R.Civ.P. 23(b)(3)) (emphasis added)).

The plaintiffs have cited several vanishing premium cases that have applied a presumption of reliance where management allegedly made material nondisclo-sures. See, e.g., In re Great S. Life Ins. Co. Sales Practices Litig., 192 F.R.D. 212, 219-20 (N.D.Tex.2000); Varacallo v. Mass. Mut. Life Ins. Co., 332 N.J.Super. 31, 752 A.2d 807, 814-18 (App.Div.2000). However, as one court in a vanishing premium case has observed on this issue, the plaintiffs here have cited no case in which reliance was presumed when individual plaintiffs did not necessarily receive the same mix of information. See Keyes, 194 F.R.D. at 257 n. 5; cf. Kramersmeier, 440 N.W.2d at 878-79 (district court ruling granting class action certification affirmed where bond purchasers alleged losses they suffered were attributable to material misrepresentations contained in the prospectus published in connection with the bond offering, and a failure to disclose certain material facts; common questions of law and fact concerning “whether the prospectus, as the common source of’ information provided by defendant, “misled the plaintiffs and other bond purchasers in alleged violation of statutory and common law”).

The complaint files here reveal the individualized nature of the sales presentations by Farm Bureau agents. For example, one client complained he purchased a second policy based on an agent’s representation that he was offering a “special policy” to a “select group.” Another agent wrote “no premiums due” on a client’s policy sheet, and sent a note congratulating the client on the purchase and stating no premiums were due. Even the facts surrounding plaintiffs’ purchase of insurance reveals the individualized nature of their claims — none of them recall receiving an illustration from their agent, Merrill Perry, before their purchase of insurance. In addition, Farm Bureau did not require agents to use materials provided by Farm Bureau for use in making sales. Therefore even assuming the nondisclosures as the plaintiffs allege, there is simply no evidence that the agents made uniform presentations, used scripts provided by Farm Bureau, and used the illustrations in all cases. Presumption of reliance would be inappropriate here because common facts, if any, do not predominate over individual facts that must be developed to determine the reliance issue.

We conclude the district court did not abuse its discretion when it (1) refused to apply a presumption of reliance to the fraud claims and (2) concluded that individual questions predominate common questions on this issue.

The district court summed up its reasons for finding that individual questions of law and fact predominate common questions this way:

In reaching this conclusion, this court is abundantly aware of its prohibition against examining the merits of plaintiffs’ claims. However, after nearly a full year of discovery, plaintiffs cannot direct this court to any piece of evidence to support their theory of uniform deception. Without more than mere speculation, this court believes that a trial as a class action would spend more time delving into individual questions of fact (■e.g. what individual agents represented to the putative plaintiffs, whether the agents employed illustrations in their sales meetings with putative plaintiffs, whether putative plaintiffs reasonably relied upon misrepresentations) than examining questions common to the entire class.

We agree with this assessment and therefore conclude the district court did not abuse its discretion when it granted Farm Bureau’s motion to decertify the class.

VII. Disposition.

Because we find no abuse of discretion in the district court’s decertification of the class action, we affirm.

AFFIRMED.