Skip to main content

Buying and cost mechanics

Does Inflation Reduce a Level Term Death Benefit? What the CPI Record Shows

Buying and cost mechanics · Updated November 4, 2026

Inflation does not change the dollar amount a level term policy pays, but it does shrink what that amount buys. A $500,000 benefit paid 20 years after a September 2006 purchase would buy what $301,092 bought at purchase, measured by the Consumer Price Index for All Urban Consumers (CPI-U).

The shrinkage is uneven across history. In the 1,124 monthly 20-year windows that began between January 1913 and September 2006, a fixed $500,000 ended with between $145,073 and $742,857 of starting-date buying power, and the median window ended at $283,165. The top of that range comes from a few years of falling prices after 1920. Of the 680 windows that began in 1950 or later, the best one ended at $334,694.

Sources and dates: how we check each figure.

What the contract does and does not adjust

A level term policy names a face amount and a term. The multistate Individual Term Life Insurance Policy Standards, adopted by the Interstate Insurance Product Regulation Commission (IIPRC) on December 15, 2022, set what a policy filed through the Interstate Insurance Product Regulation Compact must say about the payment. Section 3.G(1) reads: "The policy shall describe how the death benefit proceeds are determined and shall describe all death benefit options available under the policy." Section 3.G(3) then sets the floor: "The death benefit proceeds paid shall be at least equal to the death benefit of the policy and any riders that are payable, plus any dividend values in the policy at the time of death, less any indebtedness and plus or minus any premium adjustments with interest. If death occurs during a period for which premium has been paid, the company shall add to the policy proceeds a refund of any premium actually paid for any period beyond the end of the policy month."

Apart from a refund of premium paid beyond the month of death, the only time-based addition in that section is interest after death. Section 3.G(2)(a) says "Interest shall accrue and be payable from the date of death," and (b) sets the rate at the one "applicable to the policy for funds left on deposit or, if the company has not established a rate for funds left on deposit, at the Two Year Treasury Constant Maturity Rate as published by the Federal Reserve." Subparagraph (c) adds interest at 10% annually, beginning 31 calendar days after the latest of three listed events, until the claim is paid. That interest runs from the date of death, so it does not compensate for the years between purchase and death.

In the text of the term standards, a search for "inflation", "cost of living", "consumer price", "indexed" and "increasing" finds no match. The writer's reading is that the standard base term form is a fixed-dollar promise: the amount is set at issue and the only thing that moves after death is the interest. Riders are filed under separate IIPRC standards, and a particular policy can differ from the standard form, so the contract's own death benefit section decides.

Two other provisions bear on inflation indirectly. The first is conversion. Section 3.F(2)(f) says the conversion provision must describe "Any evidence of insurability requirements. Such evidence would be limited to amounts in excess of the amount of life insurance in effect at the time of conversion and any supplemental benefits unrelated to the life risk." The writer's reading of that wording is that converting to a permanent policy carries over the coverage already in force without new health evidence, while anything above it can be underwritten. The conversion options guide covers deadlines and credits in conversion options. The second is renewal. Section 3.V says a renewable policy "shall state the conditions, limitations or restrictions of renewal. The company may not require evidence of insurability upon renewal," and that the policy "shall contain the schedule of guaranteed premiums for the renewal periods." Renewal is described in terms of evidence and guaranteed premiums, not of a face amount that rises with prices.

Because the contract fixes the dollars and the economy moves the prices, the question for a reader is a measurement question: how much does a fixed sum lose over 10, 20 or 30 years, and how much does that loss depend on when the policy was bought? The Bureau of Labor Statistics (BLS) publishes the CPI-U monthly from January 1913, which allows 10-, 20- and 30-year windows across more than a century.

The price index and its limits

The series is the CPI-U, U.S. city average, all items, not seasonally adjusted (series CUUR0000SA0, base 1982-84 = 100). The series runs monthly from January 1913 through September 2026, a total of 1,364 monthly values (retrieval details are under Sources). The September 2026 index is 336.940. The BLS Consumer Price Index news release for September 2026 reports that the all items index rose 3.7 percent over the 12 months before seasonal adjustment, and the same index values give 3.74% (336.940 divided by 324.800 for September 2025). The release for October 2026 is scheduled for Tuesday, November 10, 2026, so the latest window ends in September 2026.

One month is missing. The BLS records October 2025 with the footnote "Data unavailable due to the 2025 lapse in appropriations." Any 10-, 20- or 30-year window that would start or end in October 2025 is left out. For the lengths used here that removes one window of each length, the one ending in October 2025, and every other window is calculated from two published values. Index values through 2006 are published to one decimal place and from January 2007 to three. The first value, January 1913, is 9.8, so a half-step of rounding on that month is about 0.5% of the index. That matters only for windows that began in the earliest years.

The BLS says plainly what the number is. In its CPI questions and answers page: "The CPI frequently is called a cost-of-living index, but it differs in important ways from a complete cost-of-living measure." The CPI-U tracks the average change over time in the prices urban consumers pay for a representative basket of goods and services. It is not any family's own experience. A household that rents, pays tuition, or buys little gasoline can see a different rate, and the section on what families pay for shows how far the components diverge.

The arithmetic is simple. If a policy pays a fixed amount F at the end of a window, the starting-date buying power of that payment is F times the CPI at the start divided by the CPI at the end. For the 1996 policy below, that is $500,000 times 157.8 divided by 336.94, which is $234,166. The windows are same-month to same-month. They overlap heavily, because neighbouring windows share all but a month at each end, so 1,124 windows of 20 years are not 1,124 independent trials, and none of the shares or medians below is a probability of anything happening to a future buyer. They describe what one record of United States prices did to a fixed sum.

Policies bought in September, year by year

The table fixes the purchase month at September and follows a fixed $500,000 to the same month 10, 20 and 30 years later. Each cell shows the sum in purchase-date dollars and the annual rate of price increase that produced it. A cell reads "Not yet reached" where the end date falls after September 2026.

Policy boughtBuying power at 10 yearsBuying power at 20 yearsBuying power at 30 years
September 1946$372,263 (2.99% a year)$311,927 (2.39% a year)$177,083 (3.52% a year)
September 1956$418,960 (1.78% a year)$237,847 (3.78% a year)$124,319 (4.75% a year)
September 1966$283,854 (5.82% a year)$148,367 (6.26% a year)$103,612 (5.39% a year)
September 1976$261,343 (6.70% a year)$182,510 (5.17% a year)$141,942 (4.29% a year)
September 1986$349,176 (3.66% a year)$271,562 (3.10% a year)$228,225 (2.65% a year)
September 1996$388,862 (2.55% a year)$326,806 (2.15% a year)$234,166 (2.56% a year)
September 2006$420,208 (1.75% a year)$301,092 (2.57% a year)Not yet reached
September 2016$358,266 (3.39% a year)Not yet reachedNot yet reached

The damage is concentrated in particular years. The policies bought in September 1966 and 1976 were the most damaged: the 1966 policy kept $148,367 of buying power at 20 years and $103,612 at 30, and the 1976 policy kept $182,510 at 20 years. The policies bought in September 1986, 1996 and 2006 did much better, with 20-year results of $271,562, $326,806 and $301,092. No 30-year cell in the table holds as much as half of the original $500,000. The best is $234,166, for the policy bought in September 1996.

The next table follows the two most recent complete cases through their terms. A death before the end of the term pays the same dollars but at a shorter distance from purchase, so the buying power is higher than at term end. The table gives each five-year mark.

Years after purchase1996 policy: dateCPI-U on that date1996 policy: $500,000 buys (September 1996 dollars)2006 policy: date2006 policy: $500,000 buys (September 2006 dollars)
0 yearsSeptember 1996157.800$500,000September 2006$500,000
5 yearsSeptember 2001178.300$442,513September 2011$447,135
10 yearsSeptember 2006202.900$388,862September 2016$420,208
15 yearsSeptember 2011226.889$347,747September 2021$369,837
20 yearsSeptember 2016241.428$326,806September 2026$301,092
25 yearsSeptember 2021274.310$287,631Not yet reachedNot yet reached
30 yearsSeptember 2026336.940$234,166Not yet reachedNot yet reached

A claim paid in the tenth year of the 1996 policy, in September 2006, delivers $500,000 worth $388,862 in September 1996 dollars, which is a 22.2% loss. Waiting until the thirtieth year takes the loss to 53.2%. Both come from the same index series. The 2006 policy is at $301,092 after exactly 20 years, which puts the loss at 39.8%. Each column loses the largest share of its value over the final five-year step shown, from $369,837 to $301,092 for the 2006 policy and from $287,631 to $234,166 for the 1996 policy. Both steps cover the same stretch, September 2021 to September 2026, when consumer prices rose 22.8% (336.940 divided by 274.310), a loss of 18.6% of the value at the start of that step in each case.

Every window since 1913

Fixed anniversaries show specific cases. The distribution of all windows shows how wide the spread is. The table below uses every monthly start from January 1913 for which the end month has a published value. The last column counts windows that ended with less than half of the original buying power, meaning below $250,000 of starting-date value.

WindowMonthly windowsLowest buying power of $500,000Median buying powerHighest buying powerWindows ending below $250,000
10 years1244$214,872 (July 1972 to July 1982; 8.81% a year)$381,599 (2.74% a year)$670,635 (May 1923 to May 1933; -2.89% a year)70 of 1244 (5.6%)
20 years1124$145,073 (January 1966 to January 1986; 6.38% a year)$283,165 (2.88% a year)$742,857 (July 1920 to July 1940; -1.96% a year)376 of 1124 (33.5%)
30 years1004$102,979 (January 1966 to January 1996; 5.41% a year)$198,049 (3.14% a year)$439,076 (June 1920 to June 1950; 0.43% a year)763 of 1004 (76.0%)

The highest results come from two stretches of falling prices, the early 1920s and the early 1930s. The 10-year window beginning in May 1923 ended with prices 25.4% lower (a multiplier of 0.746), so a fixed $500,000 was worth $670,635. The 20-year window beginning in July 1920 ended at $742,857, and the best 30-year window, beginning in June 1920, still ended at $439,076, a loss. Counting every start month, 146 ten-year windows and 58 twenty-year windows ended with an index at or below the starting index, and none of the 1,004 thirty-year windows did. The ten-year ones all began between July 1919 and August 1931, and the twenty-year ones between February 1918 and December 1922.

Falling prices of that kind did not return over a ten-year span: no 10-year window that began in 1950 or later ended with lower prices than it started with. The next table repeats the calculation for windows that began in January 1950 or later.

WindowMonthly windowsLowest buying power of $500,000Median buying powerHighest buying powerWindows ending below $250,000
10 years800$214,872 (July 1972 to July 1982; 8.81% a year)$379,508 (2.80% a year)$441,667 (January 1952 to January 1962; 1.25% a year)70 of 800 (8.8%)
20 years680$145,073 (January 1966 to January 1986; 6.38% a year)$269,083 (3.15% a year)$334,694 (January 2001 to January 2021; 2.03% a year)309 of 680 (45.4%)
30 years560$102,979 (January 1966 to January 1996; 5.41% a year)$138,020 (4.38% a year)$257,281 (January 1991 to January 2021; 2.24% a year)544 of 560 (97.1%)

Since 1950 the record is harsher. In every one of the 680 twenty-year windows that began in 1950 or later, a fixed $500,000 ended at $334,694 or less in buying power. Of those windows, 309 (45.4%) ended below half. Over 30 years, 544 of 560 windows (97.1%) ended below half, and the best of them left $257,281. The lowest cases for all three lengths sit in the 1960s and 1970s: the worst 10-year window ran from July 1972 to July 1982 at 8.81% a year, and the worst 20- and 30-year windows both began in January 1966.

The median window is the closest thing to a typical case, and the annual rate behind it is close to 3% a year. The 20-year median is 2.88% a year across all history and 3.15% for starts since 1950, and the 30-year median is 3.14% and 4.38%. The spread is the point. Over 30 years the all-history annual rates run from 0.43% to 5.41%, and a family cannot know in advance which part of that range its own term will land in.

Which start years did the damage

Grouping 20-year windows by the decade in which they began shows the shape of the record.

Start dates (20-year windows)WindowsLowestMedianHighest
January 1913 to December 191984$361,314$389,898$675,000
1920 to 1929120$348,980$485,955$742,857
1930 to 1939120$235,955$254,545$363,830
1940 to 1949120$234,899$285,039$350,000
1950 to 1959120$191,656$256,732$322,738
1960 to 1969120$145,073$148,898$188,303
1970 to 1979120$147,566$173,333$227,867
1980 to 1989120$230,450$272,593$291,967
1990 to 1999120$293,972$319,133$328,564
January 2000 to September 200680$299,795$304,248$334,694

The windows that began in the 1960s are the worst group. The median of $148,898 means that a typical 20-year term starting in that decade ended with less than 30% of its original buying power, and even the best of them, at $188,303, kept under 38%. The 1970s starts are a little better but still centre on $173,333. After the 1980s, results recover: the 1990s starts all ended between $293,972 and $328,564, a narrow band, and the windows that started from 2000 to September 2006 sit between $299,795 and $334,694. The 20-year result for the policy bought in September 2006, $301,092, is therefore in line with the recent group, and it follows two decades of inflation near 2.6% a year.

The recent group is not evidence that the problem is over. The 12 months ending in September 2026 show 3.74%, above the 20-year annual rates of the policies bought in September 1996 and 2006 (2.15% and 2.57%). The BLS data also show how quickly a quiet decade can end: the 10-year result for the policy bought in September 2016 is $358,266, a worse start than the 10-year result of the policy bought in September 2006, which was $420,208.

Hypothetical 2, 3 and 4 percent paths

The table gives the same $500,000 under three constant rates, next to the record for the window that ends in September 2026. The three rates are hypotheticals chosen for the arithmetic, not forecasts. The 2% column is not the Federal Reserve's CPI target: the Federal Reserve's page on its inflation goal says the Federal Open Market Committee "judges that inflation of 2 percent over the longer run, as measured by the annual change in the price index for personal consumption expenditures, is most consistent with the Federal Reserve's mandate for maximum employment and price stability." That is a different index from the CPI-U used everywhere else here. The page was last updated on August 22, 2025.

TermHypothetical 2% a yearHypothetical 3% a yearHypothetical 4% a yearActual CPI-U, window ending September 2026
10 years$410,174$372,047$337,782$358,266 (September 2016 to September 2026; 3.39% a year)
20 years$336,486$276,838$228,193$301,092 (September 2006 to September 2026; 2.57% a year)
30 years$276,035$205,993$154,159$234,166 (September 1996 to September 2026; 2.56% a year)

A constant 3% a year for 20 years leaves $276,838, which is close to the median 20-year window in the full record ($283,165) and above the median for starts since 1950 ($269,083). The 3% illustration is therefore a fair centre for the record. What it hides is the width around it. The worst 20-year window left $145,073, which is less than a constant 4% path would leave ($228,193), and the most recent 20-year window left $301,092, which is more than a constant 3% path. The recent 10-year window, at 3.39% a year, ended below the constant 3% path ($358,266 against $372,047).

Starting face needed to hold value

The same index values can be turned around: what face amount at purchase would have paid $500,000 of purchase-date buying power at the end of the term? The answer is $500,000 times the end index divided by the start index, which is the price multiple of the window.

TermStarting face needed: worst windowMedian windowBest windowWindow ending September 2026
10 years$1,163,484$655,137$372,781$697,806
20 years$1,723,270$882,877$336,538$830,310
30 years$2,427,673$1,262,312$569,378$1,067,617

This is arithmetic about the past, not a sizing rule. The 20-year window ending in September 2026 would have needed a starting face of $830,310. The worst 20-year window would have needed $1,723,270 and the median would have needed $882,877. What a larger face costs is on each insurer's rate pages and is not computed here. A reader holding quotes for two face amounts can line them up in the quote worksheet, which computes only what the reader enters, stores nothing and sends nothing. A larger face also has to be justified by a need that grows with prices. Whether the need does grow is the subject of the sections on what families pay for and on shrinking and rolling needs below.

What families actually pay for

The all-items index is an average. The series below are drawn from the same BLS API, with the same September-to-September windows ending in September 2026. Each cell is the multiple by which the measure grew.

MeasureBLS seriesSept 2016 to Sept 2026Sept 2006 to Sept 2026Sept 1996 to Sept 2026
All items (CPI-U)CUUR0000SA01.40 times1.66 times2.14 times
ShelterCUUR0000SAH11.49 times1.85 times2.51 times
Day care and preschoolCUUR0000SEEB031.43 times1.94 times3.10 times
College tuition and feesCUUR0000SEEB011.23 times1.92 times3.49 times
Medical careCUUR0000SAM1.27 times1.75 times2.59 times
FoodCUUR0000SAF11.42 times1.79 times2.27 times
Average hourly earnings, production and nonsupervisory employees (September 2026 preliminary)CEU05000000081.50 times1.93 times2.67 times

In the 20-year and 30-year columns, every other row rose by more than the all-items row. Day care and preschool, a cost that sits squarely inside the years a young family buys term insurance, rose 1.94 times over 20 years against 1.66 times for all items. College tuition and fees rose 1.92 times over 20 years and 3.49 times over 30. The 10-year column is more mixed: shelter (1.49), day care (1.43), food (1.42) and the earnings series (1.50) exceeded the 1.40 for all items, while tuition (1.23) and medical care (1.27) did not. The CPI rows are index levels, not dollar prices, the earnings row is dollars per hour, and the table compares only how much each series grew.

Income is the other side of the ledger. The earnings series is the BLS average hourly earnings of production and nonsupervisory employees on private payrolls, not seasonally adjusted; the September 2026 value is flagged preliminary. It grew 1.93 times over 20 years, against 1.66 times for prices. The writer's reading is that an income-replacement need tied to a paycheck has tended to grow faster than the CPI-U, at least for that group of workers, so a death benefit sized to a paycheck loses more ground than the all-items figures show. That reading is about an average and does not describe any one worker's pay.

A fixed-rate mortgage is the opposite case. The Consumer Financial Protection Bureau says: "With a fixed-rate mortgage, the interest rate is set when you take out the loan and will not change." That sentence is about the interest rate only. In the standard amortization formula used for the hypothetical below (the writer's arithmetic, not a CFPB statement), a fixed-rate loan has a constant principal-and-interest payment while the balance falls. A death benefit sized to a mortgage therefore covers a debt that shrinks in dollars while the benefit shrinks in buying power. Property taxes and homeowners insurance are not addressed by that sentence. The guide on mortgage protection versus level term sets out how a decreasing benefit follows an amortization schedule.

Shrinking needs against rolling needs

A fixed benefit loses buying power every year. Whether that loss leaves a family short depends on what the benefit is meant to cover. The hypothetical household below is invented. A $500,000 30-year term policy is bought in September 1996 by a parent whose youngest child has 15 years left to age 18. The household has a $200,000 mortgage on a 30-year fixed loan at 7.00%, a rate chosen for the arithmetic and not a market quote. The monthly principal-and-interest payment is $1,330.60. It also wants $20,000 a year, in September 1996 dollars, to supplement a surviving parent's income for the years until the youngest child turns 18. The face amount equals the need on the purchase date: $200,000 plus 15 years at $20,000.

Two need profiles follow the policy through time, both measured in nominal dollars at each mark using the actual CPI-U path. Profile A is the shrinking need just described: the mortgage balance, a fixed debt in nominal dollars, plus the remaining years of income support, priced at the price level of that date. Profile B is a rolling need, such as a standard of living a surviving spouse with no earnings must keep funding, which starts at $500,000 and moves with the CPI-U.

Years after purchaseBuying power of the $500,000 (September 1996 dollars)Mortgage balance (hypothetical)Income support still needed (hypothetical)Profile A need: balance plus income supportProfile B need: $500,000 rising with CPI-U
0 years$500,000$200,000$300,000$500,000$500,000
5 years$442,513$188,263$225,982$414,245$564,956
10 years$388,862$171,625$128,580$300,205$642,902
15 years$347,747$148,038$0$148,038$718,913
20 years$326,806$114,600$0$114,600$764,981
25 years$287,631$67,198$0$67,198$869,170
30 years$234,166$0$0$0$1,067,617

In Profile A the $500,000 face exceeds the need at every mark after the purchase date: $414,245 at five years, $300,205 at ten, $148,038 at fifteen. The loss of buying power barely matters, because the mortgage and the child-rearing years run off faster than prices rise. By the end of the term there is nothing left to cover. In Profile B the need exceeds the face at every mark after the purchase date, from $564,956 at five years to $1,067,617 at thirty, while the fixed $500,000 buys $234,166 of September 1996 dollars at the end. Most households hold some mix of the two, and the mix changes over the term. The practical test is to list each dollar need, mark whether it ends, shrinks or follows prices, and compare the face amount against the sum at several future dates rather than only at purchase. The guide on how much life insurance coverage covers the sizing side, and the chance of dying during a term puts the timing of a claim in context.

Ways to offset the loss

Education only. This site does not sell insurance and does not provide quotes, and none of the following is a recommendation. Each item is a mechanism a reader can ask a licensed life insurance agent in their state about.

A larger starting face

The table of starting faces shows the size of the adjustment that history would have required: from $336,538 in the best 20-year window to $1,723,270 in the worst. That range is wide, which is a reason to treat it as a stress test and not a target. How the cost scales with the face amount is a question for the insurer's rate pages. The comparison of term lengths shows how the number of years changes the picture.

A second policy later

Adding coverage at a later date is possible, but it is a new application at an older age and new underwriting. Whether a later purchase is approved, and in which class, depends on health and the insurer, and no page here can predict it. The guide on what underwriters ask covers the application. The riders guide lists a guaranteed insurability rider, which it describes as allowing the owner to buy "additional coverage at set future dates without new underwriting." The dates, limits and conditions are in the rider wording, and the riders guide is the place to start before asking an insurer for the form.

Conversion

The IIPRC conversion language quoted above limits evidence of insurability to amounts above the coverage in force at conversion. Conversion preserves what the family already has, in a permanent form, and extends it past the end of the term. It does not index the amount. A permanent policy with its own growth features is a different product, and the comparison of term and whole life separates the two.

Laddering matches shrinking needs, not rising prices

A ladder splits coverage into policies that end at different dates. Suppose, in the hypothetical above, the parent buys a $300,000 30-year policy and a $200,000 15-year policy instead of one $500,000 30-year policy. The total is $500,000 for the first 15 years and $300,000 afterward. In September 2011 the $300,000 policy buys $208,648 of September 1996 dollars, against $347,747 for the single $500,000. A ladder does not offset inflation. It lowers the nominal coverage on a schedule, so it fits Profile A, where the needs fall, and fits Profile B poorly, where they rise. Within a ladder, the way to add buying power is to add a new policy later, which brings the underwriting point back. The term length selection math guide works through ladders and re-buying with labelled hypotheticals.

The proceeds after payment

After a claim is paid, the lump sum belongs to the beneficiary, and what the beneficiary does with it is outside the policy. The Treasury sells inflation-linked securities, and one of them shows the mechanism. TreasuryDirect says of Treasury Inflation-Protected Securities: "The principal (called par value or face value) of a TIPS goes up with inflation and down with deflation," and "Because we pay interest on the adjusted principal, the amount of interest payment also varies." It adds, about maturity: "When a TIPS matures, you get either the increased (inflation-adjusted) price or the original principal, whichever is greater." Those sentences describe the security, not the suitability of holding it, a rate of return, or the risks of any other investment. Questions of that kind belong with a licensed financial professional. The payout options guide covers the settlement choices the insurer itself offers: life insurance payout options.

Return-of-premium term

A return-of-premium policy refunds premiums if the insured outlives the term, which is a refund fixed in dollars, subject to the same price loss. The guide on return-of-premium term shows the 2005 to 2025 version of the calculation using annual averages.

Checking your own policy

Nothing here predicts what a given insurer's form says or what a particular family should buy, and the contract controls over any summary. A short list of questions turns the history above into checks on one policy.

  1. What does the death benefit section say, word for word? Look for any language about indexing, increases or cost of living. If none appears, nothing in that section adjusts the amount.
  2. Is the benefit level or decreasing? The specifications page of a decreasing term policy shows a schedule of amounts, and a level one shows a single face amount.
  3. When does the term end, and what happens then? Compare the end date against the years in which the family's needs still run, using the future-date test from the shrinking and rolling sections.
  4. What are the conversion deadline and conditions? Check the conversion period on the cover page or specifications page.
  5. Is there a rider that allows more coverage later? If so, read its dates, limits and conditions in the rider itself. The cost of living rider is the common version, and its increase offers work differently from a larger starting face.
  6. Which needs follow prices, which end and which shrink? Day care, tuition and a surviving spouse's income are in the first group on the evidence of the table above; a fixed-rate mortgage balance is in the second.

A licensed life insurance agent in your state can answer the policy-specific parts. The hub on buying and cost mechanics collects the related guides.

Sources

  • U.S. Bureau of Labor Statistics, Public Data API (version 1, api.bls.gov): CPI-U, U.S. city average, all items, not seasonally adjusted, series CUUR0000SA0, January 1913 to September 2026; shelter (CUUR0000SAH1), college tuition and fees (CUUR0000SEEB01), day care and preschool (CUUR0000SEEB03), medical care (CUUR0000SAM) and food (CUUR0000SAF1) for the CPI-U; average hourly earnings of production and nonsupervisory employees, total private, not seasonally adjusted (CEU0500000008). Retrieved in ten-year blocks from January 1913 and checked on November 4, 2026. The October 2025 value carries the BLS footnote "Data unavailable due to the 2025 lapse in appropriations." The September 2026 earnings value is flagged preliminary. Series titles checked on the BLS series pages for CUUR0000SAH1, CUUR0000SEEB01, CUUR0000SEEB03, CUUR0000SAM, CUUR0000SAF1 and CEU0500000008.
  • U.S. Bureau of Labor Statistics, Consumer Price Index news release, September 2026 (all items index 336.940; 3.7 percent 12-month change before seasonal adjustment; October 2026 release scheduled for November 10, 2026), checked November 4, 2026.
  • U.S. Bureau of Labor Statistics, Consumer Price Index questions and answers (page last modified September 25, 2025), checked November 4, 2026.
  • Interstate Insurance Product Regulation Commission, Individual Term Life Insurance Policy Standards (sections 3.F, 3.G and 3.V, adopted December 15, 2022, in effect for new product submissions from April 3, 2023 and for new issues of previously approved products from October 3, 2023), checked November 4, 2026.
  • Board of Governors of the Federal Reserve System, Why does the Federal Reserve aim for inflation of 2 percent over the longer run? (last updated August 22, 2025), checked November 4, 2026.
  • Consumer Financial Protection Bureau, What is a fixed-rate mortgage? (last reviewed January 14, 2025), checked November 4, 2026.
  • TreasuryDirect, Treasury Inflation-Protected Securities (TIPS), checked November 4, 2026.

Questions people ask about inflation and term benefits

Does inflation reduce a level term life insurance death benefit?

The dollar amount does not change, but what it buys does. On the Consumer Price Index for All Urban Consumers, $500,000 paid in September 2026 on a policy bought in September 2006 buys what $301,092 bought at purchase. For a policy bought in September 1996 and paid after 30 years the figure is $234,166.

How much buying power does a fixed death benefit lose over 20 years?

In the 1,124 monthly 20-year windows that began between January 1913 and September 2006, a fixed $500,000 ended with between $145,073 and $742,857 of starting-date buying power, and the median was $283,165. Among the 680 windows that began in 1950 or later, the best one ended at $334,694. The windows overlap, so they describe a historical record and not a probability.

Is a 3% inflation assumption a fair illustration?

As a hypothetical it sits near the middle of the record. A constant 3% for 20 years leaves $276,838 of $500,000, against a median of $283,165 across all 20-year windows. The spread is wide: annual rates over 30-year windows ran from 0.43% to 5.41%, and the 12 months to September 2026 show 3.74%. None of the rates is a forecast.

Does converting or renewing a term policy adjust the benefit for inflation?

The Individual Term Life Insurance Policy Standards say conversion evidence of insurability is limited to amounts above the coverage in force at conversion, and that a renewable policy may not require evidence of insurability on renewal. A search of the standards' text for "inflation", "cost of living", "consumer price", "indexed" and "increasing" finds no match. Riders and individual contracts can differ, so the policy's own wording controls.

Does laddering several term policies protect against inflation?

No. A ladder splits coverage into policies that end on different dates, so the nominal amount steps down over time. That fits needs that shrink, such as a mortgage balance or the remaining years of child-rearing. In a hypothetical ladder bought in September 1996, a $300,000 policy still in force in September 2011 buys $208,648 of September 1996 dollars, against $347,747 for a single $500,000 policy.

Would a bigger face amount have offset the loss?

On the historical record, holding $500,000 of purchase-date buying power over 20 years would have needed a starting face from $336,538 in the best window to $1,723,270 in the worst, and $830,310 for the window ending September 2026. That is arithmetic about the past, not a sizing rule. Whether a larger face is worth its cost depends on whether the family's needs follow prices, which is a question for a licensed agent.