The survivor percentage you elect is priced into your very first check.

Because the insurer is obligated until the second death, it prices against the couple’s joint and last survivor life expectancy, which is longer than either individual expectancy. The distinction matters and is often blurred: joint life expectancy is how long both remain alive, which is shorter than either individual expectancy. The IRS tables name the right one on their face — Publication 939 Tables II and VI are “Ordinary Joint Life and Last Survivor Annuities.” That longer measure is why the same premium starts smaller on a joint contract than on a single life. Nothing is being taken away; the same pool is spread over a longer expected payout period.

The gap between these options is not fixed. It moves with both ages and sexes, the age difference between the annuitants, and where income pricing sits when you buy. A wide age gap makes a high survivor percentage materially more expensive, because the insurer expects a long payout after the first death. Quote the same premium across several options rather than reasoning from a percentage.

Confirm which form each carrier is quoting before you compare the numbers. At the same stated percentage a joint and contingent quote will generally look better than a joint and survivor quote, because the insurer is promising less. Carrier terminology is not standardised, so confirm the form rather than the label.

§ 401(a)(11)(A) requires that, for a vested participant who does not die before the annuity starting date, “the accrued benefit payable to such participant is provided in the form of a qualified joint and survivor annuity,” with a qualified preretirement survivor annuity for the spouse of a participant who dies before that date. Under § 401(a)(11)(B) that reaches any defined benefit plan, any defined contribution plan subject to the § 412 funding standards, and a participant under any other defined contribution plan unless conditions are met — including that the benefit is fully payable to the surviving spouse and that the participant does not elect a life annuity. That second one is a test of what the participant elects, not of what the plan offers, and it is why many 401(k) plans sit outside the rule. It is not a rule about every qualified plan, and it is not limited to married participants: under Treas. Reg. § 1.401(a)-20, Q&A-25(a), “a QJSA for a participant who is not married is an annuity for the life of the participant.”

The percentage band is statutory. § 417(b) defines a qualified joint and survivor annuity as one providing “a survivor annuity for the life of the spouse which is not less than 50 percent of (and is not greater than 100 percent of) the amount of the annuity which is payable during the joint lives of the participant and the spouse,” and which is “the actuarial equivalent of a single annuity for the life of the participant.”

The default is not the only percentage a covered plan must offer. Under § 417(a)(1)(A)(ii) a participant who waives the QJSA “may elect the qualified optional survivor annuity at any time during the applicable election period,” and § 417(g) fixes what that is: where the plan’s survivor annuity percentage “is less than 75 percent, the applicable percentage is 75 percent,” and where it “is greater than or equal to 75 percent, the applicable percentage is 50 percent.” In practice that gives a covered participant a statutory right to a second, specified survivor level rather than a single take-it-or-leave-it default.

Waiving it takes more than a signature from the participant. § 417(a)(1) lets a participant elect during the applicable election period to waive the QJSA form, but § 417(a)(2) provides that the waiver “shall not take effect unless the spouse of the participant consents in writing to such election,” and the consent must acknowledge the effect of the election and be “witnessed by a plan representative or a notary public.” There is a third requirement that is easy to miss and does real work: under § 417(a)(2)(A)(ii) the election must “designate a beneficiary (or a form of benefits) which may not be changed without spousal consent,” unless the spouse’s consent expressly permits later designations without further consent. A spouse is therefore consenting to a specific beneficiary and form, not signing a blank cheque. The statute excuses consent only where there is no spouse, the spouse cannot be located, or in other circumstances the Secretary prescribes.

None of that governs an annuity you buy with your own savings — you may choose single life, joint life, or whatever the carrier offers. But the mandate exists for a reason worth borrowing: a single life payout ends at the first death, and that is where an income cliff comes from.

For a contract bought with after-tax money, § 72(b)(1) excludes the part of each payment that bears the same ratio to the payment as the investment in the contract bears to the expected return. On a joint contract the expected return is computed on both lives. Where the payment does not change on the first death, the size of the payment cancels: it sits in the amount received and again in the expected return, so the excluded amount reduces to the investment in the contract divided by the applicable expectancy multiple. Where the survivor payment is reduced — the case this page is mostly about — it does not cancel that cleanly: the expected return is the reduced payment over the last-survivor multiple plus the excess over the joint-life multiple, so the two multiples both enter the answer. The joint and last survivor multiple is the longer one, so the excluded dollars are smaller. Treas. Reg. § 1.72-5(b)(2) sets out the computation where the survivor payment is reduced. The cancellation above is arithmetic, not something the regulation illustrates: its examples vary the mortality table rather than the payment.

One large exception belongs here rather than in a footnote. If the annuity is a QJSA from a qualified employer plan with an annuity starting date after November 18, 1996, none of the above applies to it: § 72(d)(1)(A)(i) provides that “subsection (b) shall not apply,” and the Simplified Method is used instead. One carve-out sits inside that exception: § 72(d)(1)(E) turns the Simplified Method off again where the primary annuitant has reached age 75 on the annuity starting date, unless the contract guarantees fewer than five years of payments — and the General Rule above governs there after all. That divides the investment in the contract by a fixed number of anticipated payments taken from a table of the annuitants’ combined ages — not a life expectancy, and not the joint and last survivor multiple.

The survivor keeps the same treatment on their continued payments. The two rules above are the ones people are surprised by: a long-lived survivor can outlive the basis and find the payment fully taxable, and where payments cease at death with basis unrecovered, the § 72(b)(3) deduction belongs on the annuitant’s final return and is only useful to a filer who itemizes. Our annuity taxation guide covers both cases.

Survivor percentages protect the second annuitant and nobody else. If both die early a plain joint life contract simply stops. A period-certain guarantee or a cash refund feature can be added to address that, and each one trims the starting payment further — so stack them deliberately rather than by default.