RESEARCH DOSSIER
Joint-Life and Survivorship Insurance Contracts
Abstract. Joint-life and survivorship insurance transfers mortality-contingent financial risk associated with two or more identified lives through a single insurance contract. Its defining distinction concerns the event that activates payment: first-to-die insurance pays at the earliest covered death, while last-to-die, or survivorship, insurance pays at the final covered death. The distinction changes the timing of liquidity, survival-state exposures, premium obligations, claim frequency, beneficiary outcomes, reserve patterns, and estate-planning applications. This research dossier examines the legal construction, mathematical structure, empirical mortality dependence, contract options, regulatory environment, and socioeconomic consequences of these products. It draws on insurance supervisory documents, enacted legal provisions, actuarial standards, mortality and dependence research, and reproducible hypothetical valuations. The analysis establishes that combined-life insurance should be evaluated with an explicit survivor-state process, rather than by simply adding two single-life contracts or assuming marginal mortality rates imply independent lifetimes. It also shows how a change in dependence, with marginal survival distributions held constant, can move the first-death and last-death discounted claim values in opposite directions. Financial consequences depend on which event is insured; legal consequences additionally depend on ownership, beneficiary arrangements, contractual language, and jurisdiction. The dossier offers a documented original analytical synthesis of established research and law without instituting an original pricing theory, legal doctrine, or proprietary model.
Keywords: Joint-life insurance; survivorship insurance; first-to-die insurance; second-to-die insurance; last survivor; joint survival probability; dependent mortality; copula; common shock; life-contingent valuation; insurance underwriting; life insurance contracts; beneficiary rights; estate liquidity; consumer protection; insurance solvency.
Introduction
The financial meaning of life insurance on more than one person rests on a precise question: which configuration of survival and death creates the insurer's obligation to pay? If a policy pays when either insured dies, its principal benefit becomes available upon the first covered death. If it pays when the final insured dies, a surviving insured can experience decades of continuing coverage without receiving the principal death benefit. The distinction determines how a household, business, trust, or beneficiary may use the contract. It also determines the stochastic event modeled by an actuary and the timing of projected liabilities.
The New York joint-life product outline (New York State Department of Financial Services, 2023) identifies first-to-die and survivorship products as distinct regulated forms. Dickson, Hardy, and Waters (2020) analyze the corresponding statuses in joint-life actuarial mathematics. A formal description of the insured event is therefore the necessary starting point for an international comparison of these contracts.
Four distinct but connected objects organize the investigation. The first is the contractual event, including the lives insured, the payment trigger, the benefit definition, and the policy's continuation rules. The second is the biometric event distribution, including two individual survival functions and their joint behavior. The third is the allocation of ownership, beneficiary entitlements, insurable interest, and estate consequences. The fourth is the insurer's responsibility to price, reserve, report, govern and explain the contract. Each is supported by a different combination of evidence. A statutory rule about estates cannot substitute for a mortality observation; an empirical finding about annuities cannot automatically determine the price of first-to-die insurance.
The principal related Science discipline is Life Insurance, with additional connections to actuarial science, insurance contract law, consumer protection, and insurance governance. The dossier concentrates on joint-life death-benefit contracts and uses related disciplines to interpret their operation. Its methodological approach is comparative analysis of primary law and supervisory documents, reconstruction of standard contingent-life identities, scrutiny of published empirical dependence research, and explicitly hypothetical cash-flow illustrations. The hypothetical examples are designed to test explanations; they are not quotations from actual commercial tariffs.
Definitions, Historical Foundations, and Contractual Forms
The insured-life configuration
Let T_A and T_B denote the remaining lifetimes of insured persons A and B measured from a common valuation or issue date. For a first-to-die policy, the candidate insured event occurs at T_F = min(T_A,T_B). For a last-to-die policy, the candidate event occurs at T_L = max(T_A,T_B). These times become claim triggers only when all applicable contractual conditions are satisfied, including the policy's in-force status, coverage term, and specified proof requirements. The expressions define the underlying mortality event; they do not override exclusions or lapse provisions.
A multiple-life contract may cover three or more people. Its payment event could then be the first, second, third, or final death, or another stated combination. The commonly discussed first-to-die and second-to-die forms are two-person special cases. The same underlying order-statistic mathematics can be extended to larger groups, but contract rights and regulatory filings must specify exactly which order event is insured.
The New York individual joint-life outline uses first-to-die and survivorship terminology to distinguish payment on the first insured death from payment after both insured lives have ended. Its language also illustrates how filing regulators examine cover-page identification, contractual provisions, nonforfeiture values, reinstatement, suicide clauses and split options. Naming the product is only the beginning of describing its obligations.
Joint-life insurance and joint-life annuities
A joint-life or joint-and-survivor annuity typically makes periodic payments while the specified survival condition holds, sometimes reducing its amount following the first death. A joint-life insurance death benefit instead pays when a specified death condition is fulfilled. The same survival-state probabilities enter both families, yet the financial timing, sign of the cash flows, beneficiary interests, and exposure to longevity differ. Dickson, Hardy, and Waters (2020) provide the life-contingent mathematical framework for this distinction.
The similarity in mathematical notation can obscure significant product differences. A last-survivor annuity may be most expensive when at least one insured remains alive for many years, because the insurer continues to pay benefits. A last-death insurance contract may be most expensive, in present value, when the final death happens earlier, because the insurer must pay the death benefit sooner. The economic influence of mortality dependence should therefore be derived separately for each benefit.
Ownership, insured lives, and the recipient
The persons whose deaths activate the benefit need not be the legal owners of the insurance policy. Ownership may be held by one insured, jointly, by a trust, or by another entity where allowed. Beneficiaries may also differ from insured persons and owners. Death can change ownership rights without itself causing the contracted principal death benefit to be paid. Accordingly, the questions who dies, who controls the contract, and who receives the proceeds are analytically separate.
At the first death under a conventional first-to-die policy, a single principal death benefit becomes due and the basic death coverage ordinarily terminates, subject to the actual wording and any conversion or rider provisions. At the first death under a conventional last-to-die policy, the principal payment remains contingent on the surviving insured's later death. A contractual rule may then change the premium, policy values, insurance options, or entitlement to modify the owner or beneficiary. The New York filing instructions describe treatment of post-first-death nonforfeiture provisions and split options, demonstrating why the actual policy form matters.
A life contract may be term, whole life, universal life, variable or unit-linked in its financial design, within applicable product rules. The NAIC's life insurance overview explains the general differences between familiar permanent and term structures. These underlying funding choices interact with, but are conceptually independent of, the multiple-life death trigger. A survivorship policy with flexible premiums and account charges requires a different lapse and cash-value analysis from a guaranteed-premium survivorship policy with the same two insureds.
Institutional history and insurable interest
Joint-life policies belong to the longer historical development of contracts contingent on human survival. Mathematical treatment of two-life statuses developed within actuarial life-contingency analysis, while commercial insurance markets introduced differentiated multiple-life contracts and underwriting practices. The Society of Actuaries' joint and last survivor survey studied features and practices of these products among US and Canadian insurers. Its scope supports the existence of a specialized product and underwriting domain; it does not establish the present global share of the market or a current sales ranking.
The historical distinction between insurance and wagers on lives also affects the legal context. Section 1 of the Life Assurance Act 1774 concerned insurance on lives where the interested party lacked a legally recognized insurable interest. The statute is evidence of a historic public-policy concern, not an exhaustive modern international rule. Modern rules on insurable interest, beneficiary nomination, trust ownership and corporate interests must be assessed within the legal system and date governing the contract.
Contract taxonomy and document inventory
Any rigorous case analysis records at least the insured persons, their issue ages and underwriting classes; the specified death event; coverage commencement and expiry; principal and rider benefits; premium-paying period and post-first-death premiums; guarantees; surrender values and loans; continuation and split rights; exclusions; simultaneous-death conditions; claim documentation; ownership; beneficiary arrangements; assignment; and governing law. Where the contract contains nonguaranteed crediting, participating dividends or flexible account charges, illustrations need to separate guaranteed and current assumptions.
This inventory is important for causal analysis as well as consumer comprehension. Two policies can both be called joint-life while exposing their owners to very different first-death and second-death outcomes. A comparison based solely on nominal premiums omits the distinction between one eventual death benefit and two separate eventual benefits, whether the policy covers the full expected lifetimes or a fixed term, and the cost of continuing insurance after the first death.
Joint Survival Probabilities and Dependence
Four survivor states
Let S_A(t)=P(T_A>t), S_B(t)=P(T_B>t), and S_AB(t)=P(T_A>t,T_B>t), each conditional on information available at the valuation date. S_AB is the probability both lives remain alive at time t. The four exhaustive states have probabilities:
These probabilities sum to one by inclusion and exclusion. They are valid for any admissible bivariate survival distribution and do not require independence. Each state can carry different premiums, insurance charges, ownership rights, policyholder options or expenses. In a first-to-die insurance contract, the transition from both alive to either single survivor, or to neither alive in a common fatal event, is the underlying mortality event. In a last-to-die contract, the transition to neither alive completes the underlying mortality condition.
The survival function for the first death is S_F(t)=P(T_F>t)=S_AB(t). The survival function for the last death is S_L(t)=P(T_L>t)=S_A(t)+S_B(t)-S_AB(t). Hence the cumulative probability of first death by t is 1-S_AB(t), and the cumulative probability that the last death has occurred by t is 1-S_A(t)-S_B(t)+S_AB(t). Necessarily T_F is no greater than T_L for each pair of realized lifetimes.
These identities yield a central evaluation principle. A mortality table for each individual is insufficient to determine the joint-life event distribution unless a defensible dependence assumption is supplied. Two couples with the same age, sex and underwriting-class marginals can have different probabilities of both members surviving the same interval, depending on correlated environmental, medical, behavioral or catastrophic exposure.
Independence as a benchmark
Conditional independence produces the familiar expression S_AB(t)=S_A(t)S_B(t). This is a useful benchmark but must be identified as an assumption. The SOA 2015 Valuation Basic Tables document single-life insured mortality categories, not a complete, empirically determined bivariate distribution for particular pairs. Marginal table entries should not be mistaken for measurements of dependence.
Even conditional independence may differ from unconditional independence. If two partners share an unobserved socioeconomic or environmental risk factor, individual lifetimes may appear independent after conditioning on that factor yet dependent when the factor is integrated out. Geographic proximity and exposure to the same catastrophe are further reasons to examine the joint distribution. Conversely, specific observed shared risks cannot determine a universal dependence sign or strength over all durations.
Frees, Carriere, and Valdez (1996) investigate dependent mortality in annuity valuation and demonstrate economically relevant dependence in the insured data they examined. Their findings justify empirical study of paired lifetimes; their numerical annuity effects are not transferable without recalculation to first-death or last-death death insurance. Different cash-flow triggers change how dependence affects present value.
Copulas and demographic dependence
A copula is a mathematical representation of the dependence between marginal random variables. With an appropriate survival-copula convention, S_AB(t) can be written as C(S_A(t),S_B(t)). The specified copula, its parameters, and the data used to estimate them determine whether the resulting bivariate model represents asymmetry, extreme dependence or changes with age. A single correlation coefficient does not fully describe a joint lifetime distribution, especially where tail dependence or a common shock is materially relevant.
Dufresne, Hashorva, Ratovomirija, and Toukourou (2018) examine age differences in joint lifetime modeling, using an insurer's paired-lifetime observations and examining characteristics including age gap and sex of the older partner. This supports explicit demographic differentiation when estimating joint survival. Its measured effect applies to its investigated data and applications; broader transport requires matched population and contract assumptions.
One persistent empirical challenge is selection. People who purchase individual life insurance may differ systematically from population mortality cohorts; married couples purchasing survivorship coverage may also differ from other life-insurance buyers. Within an insured pair, age, health, duration since issue and underwriting classification may be correlated. Consequently, evidence about general-population married couples or annuitants should not be presented as a direct estimate of claim rates in a modern, medically underwritten survivorship portfolio.
Common-shock mortality
Ordinary lifetime dependence and simultaneous fatal shocks are distinct modeling features. A common shock may kill both insureds at the same time, or near enough that the contract's order-of-death wording becomes decisive. Models derived from Marshall–Olkin structures are useful precisely because they admit common-shock dependence. Gobbi, Kolev, and Mulinacci (2019) examine extended Marshall–Olkin joint-life pricing, incorporating both dependence and simultaneous fatal-event mechanisms.
De Giovanni, Pirra, and Viviano (2025) extend joint mortality modeling in subordinated linear hypercube models. Their research demonstrates how dependence can be combined with dynamic mortality developments in multi-life pricing. Such methods require assumptions and validation; the presence of a sophisticated stochastic model does not establish the distribution of a particular insurer's future claims.
If both insureds die at the same instant, T_F=T_L in that realization. The last-death claim may become much earlier than an independent-lifetime model would project. But even identical biometric event times do not automatically resolve claim administration or beneficiary succession: contractual exclusions, order-of-death presumptions, proof of death and applicable law may still matter.
Bounds and identification
For any admissible joint survival model, the Fréchet bounds are
These bounds express how much the joint probability is constrained by the two marginals alone. The product S_A S_B is an independence benchmark inside the permissible region. Marginal mortality estimates by themselves do not determine a point within that region. A premium calculated with no stated dependence assumption therefore hides a material source of model uncertainty.
A credible estimation exercise uses pair-level linked experience, with dates of entry, censoring, age at issue, underwriting type, common household or catastrophic exposures, and dates of the first and second deaths. It examines selection and missingness, assesses both marginal goodness of fit and dependence, and performs sensitivity tests where second-death observations are sparse. In the absence of that data, illustrating plausible model bounds is scientifically preferable to fabricating a precise premium effect.
First-Death and Last-Death Benefit Mechanisms
Life-contingent present value
Let B(t) denote the death benefit payable at elapsed time t, v(t) the applicable discount factor, and I(t) an indicator that all necessary policy conditions remain satisfied. For the covered death time T, the actuarial present value is E[v(T)B(T)I(T)]. This notation makes explicit that death timing alone does not define payment: the policy must also remain in force and the benefit conditions must be met.
For a constant benefit B payable immediately at death, a constant force of interest r>0, eventual death, and no lapse or exclusion, the whole-life death-benefit present value is B E[e^{-rT}]. If T has density f_T, it equals B times the integral from zero to infinity of e^{-rt} f_T(t) dt. A fixed term n replaces infinity with n and pays only on events during the covered term. If the policy pays at the end of the year of death or after claim settlement, discounting changes accordingly.
For annual end-of-year-of-death benefits with discount factor v, the first-death and last-death present values are respectively
and
These are standard cash-flow identities applied to the corresponding event distributions. The annual sum truncates at the insured term for term insurance. The relevant survival basis for premium payments depends on the contract. Premiums payable only while both people are alive use an annuity-due factor based on S_AB(k); premiums continuing as long as at least one person remains alive use S_L(k); changing premiums after the first death require explicit state-dependent cash flows.
An illustrative net premium can be expressed as expected benefit present value divided by the matching premium-annuity factor. That equality omits expense, premium tax, profit margins, commissions, policyholder options, lapse, capital costs and numerous product-specific obligations. The ASOP No. 54 pricing standard (Actuarial Standards Board, 2018) provides professional context for realistic assumptions, risk assessment, sensitivity and documentation.
A reproducible independent-lifetimes example
Assume two hypothetical independent residual lifetimes, each exponentially distributed, with constant annual mortality intensities 0.010 for A and 0.015 for B, and a force of interest of 0.030. Let a unit death benefit be payable immediately upon the covered death, with no policy lapse, expenses, exclusions or mortality improvement. These are constructed mathematical parameters, not empirical insurance rates.
The first-death time is exponentially distributed with intensity 0.025, so its whole-life value is 0.025/(0.030+0.025)=0.454545, rounded to six decimals. The last-death survival function is e^{-0.010t}+e^{-0.015t}-e^{-0.025t}. Its whole-life value is
For 20-year cover, the first-death value becomes (0.025/0.055)(1-e^{-1.10}), approximately 0.3032. The last-death value becomes 0.25(1-e^{-0.80})+(1/3)(1-e^{-0.90})-(0.025/0.055)(1-e^{-1.10}), approximately 0.0322. These are expected discounted claim values; they are not quoted commercial premiums. The difference arises because the first death ordinarily occurs much earlier than the final death.
A dependence sensitivity example
Retain the same individual exponential lifetime distributions but impose a stylized perfectly positively dependent construction: define one unit-exponential variable Z and set T_A=Z/0.010 and T_B=Z/0.015. Both marginals remain unchanged. Now the first-death time has exponential intensity 0.015, producing a whole-life present value 0.015/(0.030+0.015)=0.333333. The last-death time has intensity 0.010, producing 0.010/(0.030+0.010)=0.250000.
Comparison with the independent case shows a striking reversal. Increased positive dependence, in this deliberately extreme example, reduces the discounted first-death claim value from 0.454545 to 0.333333 and increases the discounted last-death claim value from 0.128788 to 0.250000. It is a transparent boundary calculation and offers no evidence that actual insured spouses exhibit perfect dependence. It does establish the importance of defining the joint distribution and trigger before valuing a product.
The direction can be derived more generally. Keep S_A and S_B fixed. If one admissible S_AB(t) is everywhere greater than another, the first-death time is stochastically later, reducing the expectation of a positive, strictly decreasing discounted payment. Since S_L(t)=S_A(t)+S_B(t)-S_AB(t), the last-death time is then stochastically earlier, increasing its discounted payment. This result requires ordered joint survival functions over the relevant horizon and otherwise identical benefit and discount conventions. Fitted copula functions may cross, so the comparison is conditional rather than universal.
First death: immediate financial protection
A first-death policy can respond to household expenses created when either income earner or caregiver dies, or to a business obligation triggered by the first death among designated partners. Its logic is an early liquidity event. The benefit can be matched to debt, child support, estate administration, business succession or other obligations, subject to actual ownership and beneficiary terms.
The policy's efficiency cannot be inferred from comparing its face amount with two separate single-life policies. A first-death contract with one payment of B and two independent single-life whole-life policies with one payment of B each create different numbers and timings of claims. The combined single-life arrangement may ultimately pay twice. An appropriate comparison states the same event needs, total available benefits, insured term, guarantees and payment pattern before evaluating relative price.
Last death: deferred financial protection
A survivorship contract links the main death benefit to the end of the last insured lifetime. This can support a planned transfer of wealth after both insureds die, provide liquidity for beneficiaries at a final-succession event, or support other end-of-life obligations. It cannot be assumed to fund immediate expenses following the first death, since its principal insured event has not then occurred.
Premium continuation is central. The surviving insured may face reduced income, health impairment or increased household expenses. If continued payments remain due, lapse risk can rise precisely when the residual death-benefit guarantee remains valuable. Split options, loans, waiver provisions, nonforfeiture values or a first-death rider can change this interaction. The New York policy outline documents variants of policy-value treatment and split options in its filing context.
Lapse, behavior and guarantees
A realistic cash-flow process includes an in-force state and potential lapse, surrender, loan, conversion or policy split events. These interact with mortality rather than occurring in a separate economic universe. After a first death, the survivor's financial capacity to pay, health, knowledge of the policy and valuation of the remaining protection can change. Premiums may be guaranteed or adjustable, while nonguaranteed charges and crediting may evolve over time.
The ASOP No. 2 standard (Actuarial Standards Board, 2021, effective 2022) addresses the development and review of nonguaranteed elements within its US professional scope. The need for disciplined assumptions applies especially where illustrations show a possible premium path rather than a fully guaranteed one. A scientifically credible model states the contract's guaranteed minimum and possible future policyholder obligations separately.
Household, Estate, and Institutional Uses
Financial timing and household needs
The economic use of joint-life coverage follows the time profile of the need being funded. The first loss of a household contributor may generate immediate obligations: mortgage payments, childcare, replacement caregiving, funeral expenses, loss of employment-linked benefits and transition costs. Last-death liquidity may instead support bequests, final estate distribution or the protection of assets intended for heirs. Other assets, benefits and insurance policies can offset these needs; there is no universal benefit amount or universally superior product.
Two policyholders who appear demographically similar may have very different contingent needs. A couple with dependents and outstanding debt has a different first-death exposure from a retired couple whose remaining assets are intended primarily for eventual beneficiaries. A household with an illiquid family business may require first-death business continuation and later succession liquidity in separate amounts and at separate times. The product event must be matched to the liability event.
Ownership, trusts and designated beneficiaries
A survivorship designation does not identify who owns the policy. Ownership may determine powers to borrow, assign, surrender, change beneficiaries, alter funding or transfer the policy to a trust. A beneficiary designation determines who receives proceeds subject to the governing law and valid contractual elections. Those choices can be modified or constrained by legal obligations, including trust terms, creditor claims and family-property rules.
A useful analytic distinction is between three events: the death of an insured, a change in ownership or legal interest upon death, and the insurer's actual claim payment. These may occur at different times and have different tax or legal consequences. An assertion that a last-to-die policy avoids all first-death estate consequences would conflate those events.
United States federal estate tax
Internal Revenue Code Section 2042 concerns proceeds of life insurance on a decedent's life receivable by the estate, or receivable by others where the decedent possessed legally specified incidents of ownership. The IRS instructions for Form 706 discuss relevant rights, such as changing the beneficiary, assigning or surrendering insurance, and borrowing against a policy. Other estate tax provisions may also be relevant to property interests, assignments or trusts.
The consequences for a two-life contract require analysis at each relevant death. A first death under a survivorship policy normally causes no principal last-death payout, yet the deceased may have held property interests or powers in the policy. At the final death, both the proceeds and the applicable ownership tests must be examined. Contract terminology cannot establish an automatic exemption. The policy's issuing jurisdiction, owner, trust documents, deceased insured, beneficiary, applicable tax year and statutory provisions are necessary facts.
United Kingdom inheritance tax
HM Revenue & Customs' joint-life policy guidance, updated April 2026, distinguishes forms of beneficial ownership, including joint tenancy and tenancy in common, when considering inheritance-tax consequences of jointly held insurance. Its treatment shows that a first death can change or transfer a legally cognizable interest even when the insurance event and payment date require separate analysis.
The US and UK examples illustrate a general comparative principle. The same pair of lives and the same death-benefit trigger can produce different legal consequences because ownership rules, estate inclusion, exemptions, trust treatment and tax instruments differ. A mathematical identity about S_AB(t) is portable across jurisdictions; the legal valuation and transfer of an interest must be established from applicable law.
Commercial and non-spousal applications
Joint-life first-death insurance can address obligations in a partnership, family company or other commercial enterprise that arise when the first designated participant dies. Whether one policy or several individual policies better serves the obligation depends on the buy-sell agreement, creditor positions, insurable interest, ownership, beneficiary requirements and policy availability. Coverage on unrelated insureds is particularly dependent on jurisdiction-specific rules and product underwriting.
The Life Assurance Act 1774 remains an important historical reference for the underlying concern with financial interest in the life insured. It supplies neither the full present-day UK rule nor a universal condition for every nation. Product and legal research must identify the current local requirements before attributing a particular structure to lawful market practice.
Affordability and beneficiary consequences
After the first insured death, the survivor may face new liquidity constraints at precisely the moment the survivorship policy needs continued premiums. This can create a distributive tension between the intended eventual beneficiaries and the current needs of the survivor. The policy may be surrendered, lapse or be borrowed against, potentially reducing the intended later proceeds. Long-term protection therefore depends on both the modeled mortality process and the social and financial circumstances of premium payers.
Consumer-facing explanations should describe observable states rather than rely only on promotional names. For each state—both insureds alive, one alive, neither alive, policy surrendered and policy lapsed—an explanation should identify whether a claim is due, who controls the policy, whether premiums continue and what guarantees survive. This is a presentation of existing contract mechanics; it does not propose a novel actuarial theory.
Comparative Regulation, Valuation, and Consumer Protection
State insurance regulation and policy terms
New York's individual joint-life filing outline supplies a concrete, dated regulatory example. It requires accurate identification of the joint product, pays attention to death-benefit definitions and addresses contestability, reinstatement, suicide clauses, split options and nonforfeiture values. The source is limited to the regulator's state filing context and may be affected by later statutory or administrative changes.
A claim dispute can arise even when all parties agree which person died first. Did coverage begin before the death? Was a premium grace period applicable? Does an exclusion apply to one insured, the first death or both? Does a split option require a deadline? Who can exercise policy options after the first death? Does a simultaneous-death provision alter the beneficiary order? These questions require the actual policy and applicable legal framework, rather than a generic statement about all survivorship products.
Distribution and informed selection in the European Union
Directive (EU) 2016/97 on insurance distribution includes demands-and-needs and customer-information obligations within its regulatory scope. In the multiple-life setting, demands-and-needs analysis ought to connect the timing of the insured event to the customer's stated contingency. A first-death family-income need and a final estate-transfer need differ even where the expected monetary claim value is similar.
Application of the directive depends on its operative provisions, national transposition and the particular product. Products containing investments may entail additional obligations. The legal text cannot establish that every actual distribution in a given market complied with the law. It does establish a framework for evaluating whether the policy's benefit trigger, term, cost and continuing obligations were communicated and matched to the identified needs.
Pricing and insurance contract measurement
The purpose of pricing is to establish a commercially and prudentially supportable premium and contract design. Professional guidance such as ASOP No. 54 directs attention to mortality, expense, investment, lapse, policyholder-option, profit and risk assumptions. A two-life model must explicitly represent the survival state and the coverage trigger. The simplified net premium identity cannot serve as a complete product tariff.
IFRS 17 Insurance Contracts provides a financial-reporting framework for insurance contract groups, including fulfillment cash flows, a risk adjustment for nonfinancial risk, and where applicable a contractual service margin. The first- or last-death obligation enters the prospective cash flows according to the contract's terms and survival probabilities. The reporting result depends on the appropriate aggregation, contract boundary, discount rates, insurance service and measurement approach, not solely the mathematical expected benefit.
The European Solvency II directive, including Article 77, defines technical provisions through a best estimate and risk margin framework. Best-estimate cash flows must reflect the insurer's future obligations and a probability-weighted projection. In a multiple-life contract, the projection should capture mortality states, premiums, expenses, guarantees, lapse and relevant other obligations. It may also need the impact of reinsurance, concentration and common mortality shocks on capital and liquidity.
Solvency and the timing of risk
First-death claims tend to arise earlier than last-death claims with the same pair of lives and contractual survival horizon. Earlier expected payment can increase liquidity demands, while later claims can increase exposure to discount-rate, long-duration guarantee and longevity assumptions. Common-fatal-shock risks may accelerate both types; dependence can alter the difference between them. These are directional mechanisms, not measured comparative capital ratios for all portfolios.
A financial model with a plausible expected claim value can still fail to represent extreme losses, surrender or funding shocks. A portfolio with many couples exposed to a shared catastrophe should be examined for correlated claims. At the same time, dependencies across insured pairs need not be the same as dependencies within each pair. Product-level two-life modeling and portfolio-level accumulation are separate layers of aggregation.
Interpretive and consumer safeguards
Material disclosures for a survivorship policy should identify the fact that the principal death benefit is ordinarily unavailable after the first death; the required funding after that event; guarantees and nonguaranteed charges; the treatment of cash value and loans; the effect of lapse; any split or conversion rights; and ownership and beneficiary procedures. For a first-death policy they should identify that the covered event is the first death, ordinarily producing one principal claim rather than independent benefits at both deaths.
These descriptions must be specific to the contract, not generalized from an industry label. Insured individuals can differ in health or age, and one person may be excluded from certain benefits. A robust consumer explanation should also distinguish tax discussion from an assurance of tax treatment. The insurer's conduct obligation is to communicate accurate policy features within the operative legal framework.
Empirical Evidence, Sensitivity Analysis, and Scientific Synthesis
The research evidence and its transfer limits
The available academic literature establishes that paired lifetimes can be dependent and that dependence affects valuations. Frees, Carriere, and Valdez (1996) studied paired annuitant mortality. Dufresne and colleagues (2018) examined spousal age-gap dependence. Gobbi and colleagues (2019) studied common-shock joint-life models. De Giovanni and colleagues (2025) investigated dynamic joint mortality structures. They investigate distinct methods and populations rather than providing one universal mortality dependence coefficient.
The present dossier does not independently analyze confidential insurer experience, calibrate a product tariff, reproduce proprietary model software or meta-analyze all empirical datasets. Its formulas follow standard probability and valuation mathematics, and its numerical results are calculated for openly specified assumptions. Journal bibliographic or abstract access is distinguished from full-text verification. This scope prevents the incorrect transformation of an annuity dependence effect into a claimed discount for life insurance.
Comparative product experiment
A reproducible comparison begins by holding the insured lives, valuation date, marginal mortality, benefit amount, discount convention and contractual term fixed. The next step specifies the first- versus last-death payment event. Separate scenarios then vary dependence, premium-paying period, lapse, guarantees, policy ownership and riders. The output should show the discounted benefit value, the distribution of the claim date, the required survivor-state premiums, the recipient and the obligations being funded.
The first-death and last-death formulas should be evaluated under an explicit independence benchmark and at least one plausible alternative dependence specification whenever data or model risk justifies it. If pair-level evidence is unavailable, present reasonable bounds or conditional stresses, not a fictitious best estimate. The SOA insured mortality tables can serve as one source for marginal insured mortality, subject to selection and version, but cannot establish pair dependence by themselves.
The qualitative impact of mortality improvement also differs across triggers. Improving longevity in one insured may leave the first-death distribution heavily affected by the other insured while extending the last-death distribution. Dependence dynamics, changes in cohort experience and catastrophic risks can complicate that result. A credible calibration and validation design states the issue population, mortality base year, projection method, age and underwriting differentiation, and the treatment of shared exposures.
Stress tests and adverse states
Mortality stresses should separately address elevated mortality for A only; for B only; for both independently; and for both through a common shock. The last case can cause the last-death claim to accelerate dramatically relative to a benchmark in which one partner survives for many years. A longevity improvement stress can have a different effect and may be most influential for a last-death guarantee at long duration. An interest-rate stress changes the discounted value of the cash flows and can also affect investment account crediting and guarantees.
Behavioral stresses should include increased lapse after first death, policy loans funding the survivor's expenses, decreased ability to pay a flexible premium, and exercise of available options. The distribution of product outcomes under these stresses can differ from the simple mortality-only expected present value. ASOP No. 54 provides a relevant practice-standard rationale for sensitivity and risk assessment, although it does not prescribe one universal stress grid for every jurisdiction.
A claims and legal stress test asks who has legal control of the policy after the first death, whether a beneficiary dies before payment, and how simultaneous death or disputed chronology is determined. A tax stress test compares cases in which the insured is also the owner, a trust has specified ownership, or another person has legal powers. These scenarios may change the legal outcome even where the insured mortality distribution stays identical.
Findings from original analytical synthesis
The first finding is the primacy of the insured event. First-to-die and last-to-die policies on identical lives can produce markedly different liquidity schedules even with the same face amount and discount basis. A comparison that omits the insured event is incomplete.
The second finding is the importance of four explicit survivor states. A policy's first-death and last-death values depend on S_AB and S_A+S_B-S_AB respectively. Marginal mortality tables do not uniquely determine either one without an assumption about dependence. Under ordered joint survival and fixed marginals, increasing joint survival across time decreases the first-death discounted claim value while increasing the last-death value. This is an analytical consequence of established probability identities, not a new theory.
The third finding concerns the boundary between actuarial and legal reasoning. A death-benefit event identifies when payment is contractually due. Estate ownership, inheritance, trust obligations and tax inclusion identify what legal consequences occur. US federal estate inclusion and UK joint-policy inheritance guidance operate in different legal frameworks, but both require analysis of legal interests rather than relying on the word survivorship.
The fourth finding concerns the relationship between consumer purpose and institutional risk. The policyholder needs protection aligned to a financial need and an intelligible description of when it pays. The insurer must project the same contract's future states to price benefits, manage liquidity, recognize liabilities and govern capital. A policy can have an actuarially supportable premium but be mismatched to the timing of the customer's need. Economic value and coverage appropriateness therefore need separate assessments.
Research agenda and evidentiary boundaries
Future empirical work should use linked experience of insured pairs, with documented issue dates, ages, underwriting categories, survivor-state transitions, policy options, actual premiums, lapse, deaths and claims. The effect of demographic differences and shared hazards should be estimated with transparent censoring, left truncation and selection controls. Model comparisons should report out-of-sample calibration and separate first-death from last-death valuation error. Legal comparisons should link sampled policies to dated jurisdictional law, actual ownership and beneficiary arrangements.
The present study draws a careful boundary around its conclusions. The numerical illustration demonstrates a sensitivity mechanism, not current market premiums. The empirical papers establish modeling concerns within their stated populations, not one universal product coefficient. The legal sources establish dated national or regional propositions, not a single international estate doctrine. External peer review has not been requested, and no proprietary model or novel theory is asserted.
Conclusion
Joint-life and survivorship life insurance forms a distinct scientific subject because its financial obligation depends on the survival configuration of several identified lives. First-to-die insurance transfers risk at the earliest covered death, typically creating an earlier benefit and ordinarily a single payment. Last-to-die insurance transfers risk at the final covered death, continuing through the first death and making post-first-death contractual obligations and survivor circumstances especially important.
The mathematical object is the joint distribution of the insured lifetimes. A four-state representation establishes each event's probability and the distribution of its payment time. The dependence structure links the marginal survival probabilities to the first- and last-death cash flows. The examined scholarly literature confirms the relevance of dependent mortality and provides statistical and common-shock modeling approaches. Independent exponential examples reproduce the substantial differences that a change in dependence can generate under fixed marginals.
The institutional and legal object requires its own evidence. Ownership, beneficiary nomination, insurable interest, premium continuation, claim exclusions, first-death options and financial reporting cannot be inferred from the survival equations alone. US tax provisions, UK inheritance-tax guidance, New York insurance filing rules, EU distribution legislation, IFRS accounting and EU solvency supervision illustrate distinct aspects of a contract that may share the same mathematical event.
The integrated conclusion is that research, pricing and legal evaluation should begin with the exact insured event and policy state; proceed through an explicit joint survival distribution, premium and benefit cash-flow model; and then examine legal and regulatory consequences separately at each relevant death. The scientific contribution of this dossier is a source-identified, internationally scoped synthesis and a transparent reconstruction of comparative actuarial mechanisms. It establishes no new actuarial theory or proprietary doctrine.
References
Actuarial Standards Board. (2018). Actuarial Standard of Practice No. 54: Pricing of Life Insurance and Annuity Products. https://www.actuarialstandardsboard.org/asops/pricing-of-life-insurance-and-annuity-products/
Actuarial Standards Board. (2021). Actuarial Standard of Practice No. 2: Nonguaranteed Elements for Life Insurance and Annuity Products (effective June 1, 2022). https://www.actuarialstandardsboard.org/asops/asop-no-2-nonguaranteed-elements-for-life-insurance-and-annuity-products/
De Giovanni, D., Pirra, M., & Viviano, F. (2025). Joint mortality models based on subordinated linear hypercubes. ASTIN Bulletin: The Journal of the IAA. https://www.cambridge.org/core/journals/astin-bulletin-journal-of-the-iaa/article/joint-mortality-models-based-on-subordinated-linear-hypercubes/78D8B42B450874CCEB8F98FEDE253D3C
Dickson, D. C. M., Hardy, M. R., & Waters, H. R. (2020). Actuarial Mathematics for Life Contingent Risks (2nd ed.), chapter 10: Joint Life and Last Survivor Benefits. Cambridge University Press. https://www.cambridge.org/highereducation/books/actuarial-mathematics-for-life-contingent-risks/281DA4E8D523A6B23280ADC3D165AFDA/joint-life-and-last-survivor-benefits/EA96623502E1FE4D19FB3EED6178B2F9
Dufresne, F., Hashorva, E., Ratovomirija, G., & Toukourou, Y. (2018). On age difference in joint lifetime modelling with life insurance annuity applications. Annals of Actuarial Science. https://www.cambridge.org/core/journals/annals-of-actuarial-science/article/on-age-difference-in-joint-lifetime-modelling-with-life-insurance-annuity-applications/06A58B152B542F306663080AB67E4F03
European Parliament and Council. (2009). Directive 2009/138/EC on the Taking-Up and Pursuit of the Business of Insurance and Reinsurance (Solvency II), Article 77. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32009L0138
European Parliament and Council. (2016). Directive (EU) 2016/97 on Insurance Distribution. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex%3A32016L0097
Frees, E. W., Carriere, J., & Valdez, E. A. (1996). Annuity valuation with dependent mortality. Journal of Risk and Insurance, 63(2), 229–261. https://doi.org/10.2307/253744
Gobbi, F., Kolev, N., & Mulinacci, S. (2019). Joint life insurance pricing using extended Marshall–Olkin models. ASTIN Bulletin: The Journal of the IAA. https://www.cambridge.org/core/journals/astin-bulletin-journal-of-the-iaa/article/abs/joint-life-insurance-pricing-using-extended-marshallolkin-models/247863E0F8E0A63471C20B133BFE6545
HM Revenue & Customs. (2026). Inheritance Tax Manual IHTM20303: Joint life and joint name policies, updated April 2026. https://www.gov.uk/hmrc-internal-manuals/inheritance-tax-manual/ihtm20303
International Accounting Standards Board / IFRS Foundation. (2023). IFRS 17 Insurance Contracts. https://www.ifrs.org/issued-standards/list-of-standards/ifrs-17-insurance-contracts/
Internal Revenue Service. (2026). Instructions for Form 706, Schedule D: Insurance on the Decedent's Life. https://www.irs.gov/instructions/i706
National Association of Insurance Commissioners. (2025). Life Insurance: Insurance Topics. https://content.naic.org/insurance-topics/life-insurance
New York State Department of Financial Services. (2023). Individual Joint Life Insurance Product Outline, revised December 19, 2023. https://www.dfs.ny.gov/system/files/documents/2023/12/Joint_Outline_2023-12-19.pdf
Parliament of Great Britain. (1774). Life Assurance Act 1774, 14 Geo. 3 c. 48, section 1. https://www.legislation.gov.uk/apgb/Geo3/14/48/section/1?view=plain
Society of Actuaries. (2015a). Joint and Last Survivor Survey Report. https://www.soa.org/resources/experience-studies/2015/joint-and-last-survivor-survey/
Society of Actuaries. (2015b). 2015 Valuation Basic Tables and Report. https://www.soa.org/resources/experience-studies/2015/2015-valuation-basic-tables
United States Congress. (2026). Internal Revenue Code, 26 U.S.C. § 2042, Proceeds of Life Insurance. https://uscode.house.gov/us/usc/t26/s2042/