The lifetime cost of mandatory tax filing to one American

Written: 2026-07-24. Status: draft, all arithmetic shown.

Quantitative backing for the April project hero framing. Source of record for the per-return burden figures is ../../docs/research/policy-03-tax-filing.md; where this document departs from that doc, the reason is stated and the research doc needs correcting (see §10).

Everything here is a floor. All IRS/NTU/Tax Foundation burden figures are federal only. Most Americans also file a state return (41 states tax individual income). No sourced state-filing burden estimate is used here, so every hour and every dollar below understates the real total.

---

Bottom line

Six numbers. Each is either a government figure or a stated arithmetic derivation from one.

Number 6 is the one that makes the hero's "spend $10 to keep her out" literally true rather than rhetorical: the IRS's own broad-scope cost estimate for a mature Direct File is $9.96 per user per year, against $290 of private cost per return.

The "two months" claim: defensible, but only for one archetype under one conversion. See §2.4 for the exact conditions. Short version: it is true for a business-income filer (29% of filers) using the IRS's tax-year-2024 estimate, under a strict 24-hour-day conversion — 1,440 hours is exactly 60 days. It is not true for the average filer, who lands at 32.5 days. Lead with "a month of your life"; use "two months" only when speaking about the self-employed, and name the tax year.

---

0. The source figures, and exactly what each one is

Everything downstream comes from these. Nothing else is used.

0.1 Per-return burden (the load-bearing numbers)

NTU Foundation, "Individual Income Tax Compliance Burdens Ease for the New Filing Season," published 2026-02-02, Table 2 — compiled from the IRS's own taxpayer burden model (the estimates the IRS files under the Paperwork Reduction Act and prints in the Form 1040 instructions):

Source: https://www.ntu.org/foundation/detail/individual-income-tax-compliance-burdens-ease-for-the-new-filing-season

Internal consistency check (do this yourself before quoting — it is what makes the table trustworthy):

• TY2024 hours: 0.71 × 8 + 0.29 × 24 = 5.68 + 6.96 = 12.64 → rounds to 13 ✓
• TY2024 dollars: 0.71 × 160 + 0.29 × 620 = 113.60 + 179.80 = $293.40 → rounds to $290 ✓
• TY2025 hours: 0.71 × 8 + 0.29 × 21 = 5.68 + 6.09 = 11.77 → rounds to 12 ✓
• TY2025 dollars: 0.71 × 160 + 0.29 × 610 = 113.60 + 176.90 = $290.50 → rounds to $290 ✓

The published headline averages are reproducible from the published components. Good sign.

Tax Foundation, working independently from the same IRS/OMB data, reports the same per-1040 split for the earlier vintage: "13 hours overall (9 hours without business income; 24 hours with business income)" (https://taxfoundation.org/data/all/federal/irs-tax-compliance-costs/). The nonbusiness figure has been reported as both 8 and 9 hours across vintages. We use 8 — it is the lower number, it is the current one, and it comes with a dollar figure attached.

0.2 Aggregate burden — and the distinction that matters most

There are two different aggregates in circulation and conflating them is the single easiest way to get fact-checked into oblivion.

The 7.1 billion / 7.9 billion hour figures are not the cost of making individuals file their taxes. They are the cost of all federal tax compliance, most of which is business and corporate. The number that belongs on a site about pre-filled 1040s is ~2.1 billion hours and ~$140 billion per year.

Cross-check: Tax Foundation independently reports "Individual Income Tax Return Hours: 2.249 billion hours, $145.4 billion" for the 2023 vintage — which is exactly NTU's Table 1 row for 2023 (2.249 bn hours, $142.32 bn). Two organizations, same number. ✓

Second cross-check: 2.129 bn hours ÷ 13 hours per return = 163.8 million returns, which is the right order for US individual return volume (~160–163M). ✓

0.3 Wage rates for valuing time

Useful credibility point: the two analyst shops on opposite sides of this fight value the hour within 17 cents of each other — $44.50 (Tax Foundation) and $44.67 (NTU, which opposes Direct File). That is not a contested number.

Verify NTU's arithmetic: 2.129 bn hrs × $44.67 = $95.1 bn, which is exactly NTU's Table 1 "opportunity cost of time burden" for 2024. ✓

0.4 Sources with a health warning

• $5.6M in prep fees saved by 140,803 Direct File pilot users (IRS, April 2024). Used below as a measured floor.
• $23 bn/year full-scale savings — Economic Security Project via Sen. Warren's office. Advocacy estimate. Do not lead with it.
• 42–48% of returns accurately pre-populatable — Goodman, Lim, Sacerdote & Whitten (Treasury economists), NBER w30008 / NTJ 2023. Academic. Used in §7 as the honest ceiling on what automation recovers.
• $64M–$249M/yr mature Direct File cost, 5M–25M users — IRS Report to Congress, May 2023, via CRS IF12654.

---

1. Lifetime hours

1.1 The filing-life assumption

60 filings, one per year from age 20 through age 79. This is the assumption the hero framing uses, and it is an upper-ish bound: people below the filing threshold do not file, and some retirees stop. Sensitivity is in §1.3.

1.2 Three archetypes

Arithmetic: 8 × 60 = 480. 13 × 60 = 780. 24 × 60 = 1,440. 12 × 60 = 720. 21 × 60 = 1,260.

1.3 Sensitivity to the number of filings

Because "nobody files for 60 years" is a real objection.

Even a 40-year filing life costs the average filer three weeks of round-the-clock life. The claim does not depend on the aggressive assumption.

1.4 A labeled lifecycle blend (illustrative only)

Most people do not stay in one archetype. A modeled path — simple in your twenties, average in your working middle, simple again in retirement:

• Ages 20–29 (10 filings) at 8 hrs = 80 hrs
• Ages 30–69 (40 filings) at 13 hrs = 520 hrs
• Ages 70–79 (10 filings) at 8 hrs = 80 hrs
• Total = 680 hours = 28.3 days

This is our model, not a published figure. It sits between the simple and average archetypes, as you would expect. Do not put it on the homepage; it is here so nobody can claim the blend would change the picture.

---

2. Converting hours into an intuitive unit

This is the section a hostile reporter attacks first, so it is written to be attacked.

2.1 The problem, stated plainly

A jail sentence is 24 hours a day of confinement. Filing is not. Converting "480 hours of filing" into "20 days in jail" is not apples-to-apples in the ordinary sense: 480 hours of filing is spread across 60 tax seasons in 8-hour chunks, and the filer sleeps in her own bed. Anyone who says "480 hours is 60 workdays, which is three months of work, which is like three months in jail" has silently switched units mid-sentence and should be called on it.

The honest move is to pick the conversion that makes the unit identical to jail — the 24-hour day — even though it produces the smallest number. It produces the smallest number because it is the strictest, and the strictest number is the one that survives.

2.2 Three honest conversions

Arithmetic: divide by 24, 16, and 8 respectively. Work-weeks = workdays ÷ 5.

In calendar months (30.44 days average):

2.3 Which conversion the site should use — and why

Use (a), the continuous 24-hour day, wherever the word "jail," "cell," or "sentence" appears. Three reasons:

1. It is the only conversion where the unit on both sides of the comparison means the same thing. A judge who says "sixty days" means 1,440 hours of custody; we mean 1,440 hours of filing. Same number, same unit.
2. It is the most conservative of the three. If a reporter tries to knock the number down, there is nowhere to knock it to — we already used the divisor most unfavorable to us.
3. It removes the only real gotcha. The moment we say "480 hours is 60 workdays, basically three months," we have handed over the story.

Use (b), the 16-hour waking day, when the sentence is about a life rather than a sentence — "48 days of your waking life." It is defensible on its own terms (nobody loses sleep hours to TurboTax) but do not pair it with the word jail.

Never use (c) in the same sentence as a jail comparison. It is the right unit for "what this would cost an employer" and nothing else.

There is a legitimate rhetorical point available and it should be labeled as rhetoric, not arithmetic: 780 hours of compelled labor is arguably a worse thing to take from a person than 780 hours of idle confinement. That is an argument about the quality of the hours, not about how many there are. Make it as an argument. Never fold it into the math.

2.4 So is "two months" true?

Yes, under exactly these conditions, and no others:

The filer has business income (a pass-through reported on the 1040 — 29% of all filers), files 60 returns, and we use the IRS's tax-year-2024 per-return estimate of 24 hours, converted at 24 hours per day.

24 × 60 = 1,440 hours ÷ 24 = exactly 60 days = 1.97 calendar months.

Where it fails:

• The average filer does not get there. 780 hours = 32.5 days = 1.07 months. Just over one month, not two. Claiming two months for the average filer is false and will be caught.
• The simple filer is nowhere near. 480 hours = 20 days.
• The TY2025 update breaks it even for the business filer. The IRS's newest estimate drops business filers from 24 hours to 21 — 1,260 hours = 52.5 days = 1.7 months. If we quote "two months," we must say "tax year 2024."
• It is not two months for one average person under any of the three conversions in §2.2, unless you switch to working days, at which point it becomes 4.5 working months, which is a different unit and must be labeled as such.

One additional true path to two months, but it costs the "your daughter" framing: two never-married adults each filing an average return for 60 years jointly consume 120 × 13 = 1,560 hours = 65 days = 2.1 months. That is two people's lives, not one. See §5.

2.5 Recommended hero copy

The strongest true version for one person:

A month of your life. Day and night. The average American spends 13 hours a year filing a federal tax return. Over a filing life, that is 780 hours — 32.5 days, counted the way a jail sentence is counted. If you are self-employed, it is two months.

"A month of your life" holds under both published vintages (32.5 days at 13 hrs/yr, 30.0 days at 12 hrs/yr) and under the strictest conversion. Nothing in it can be knocked down.

---

3. Lifetime dollars

3.1 Out-of-pocket preparation cost, 60 filings

3.2 Nominal vs. real — say this before anyone asks

60 × $290 = $17,400 is a sum of constant 2024 dollars, not a nominal lifetime outlay. That is the correct way to state a lifetime real cost, and it is the honest one.

A nominal sum — inflating each year's $290 forward — would produce a much bigger and completely meaningless number (dollars of 2026 and dollars of 2086 are not the same unit and must not be added). Do not produce that number. If asked, the answer is: "$17,400 in today's dollars, which is the only version that means anything."

The assumption embedded in $17,400 is that the real per-return cost stays roughly flat. It has not been flat — NTU's Table 1 shows individual out-of-pocket costs rising from $31.7 bn (2017) to $49.8 bn (2025), and the average per return rising from $270 to $290 between the TY2023 and TY2024 estimates. Flat is therefore the conservative assumption. Say so.

3.3 Opportunity cost: what the fees would be worth invested

Our assumption, clearly labeled: a 5% real (after-inflation) annual return, roughly a diversified 60/40 portfolio's long-run real return. Bands at 3% and 7% shown so the reader can pick their own.

Because the return is real and the contributions are constant real dollars, the future value is also in today's dollars. This matters: $102,500 at age 80 means $102,500 of today's purchasing power, not an inflated nominal figure.

Method: ordinary annuity (payment at the end of each year — the conservative timing), n = 60.

FV = P × [(1+r)^60 − 1] / r

Annual fee invested instead:

Arithmetic (base case): 160 × 353.584 = 56,573; 290 × 353.584 = 102,539; 620 × 353.584 = 219,222.

Headline: the average filer's lifetime tax-prep fees, invested instead, are about $100,000 in today's dollars at age 80. Roughly six times the $17,400 of fees themselves.

3.4 Full economic cost invested (label carefully)

If you also monetize the hours at the median wage and invest that too:

Flag this hard. It assumes those hours would otherwise have been converted into wages and invested, which is not literally true for most people (you cannot sell a Saturday afternoon in March at the median wage). Use §3.3's fee-only figure as the headline; this table is for the "full economic cost" footnote only.

---

4. The value of the lifetime hours

Four rates, three archetypes, 60 filings, TY2024 figures.

Arithmetic: 780 × 7.25 = 5,655; 780 × 24.51 = 19,117.80; 780 × 44.50 = 34,710; 1,440 × 56.67 = 81,604.80.

The spread is 7.8×, from $7.25 to $56.67. That is a lot of room for a hostile reader to move the number, so state the rate every single time.

Which rate to use, and why:

• Use $24.51 (BLS median hourly wage) for claims about an individual. It is a direct government statistic, it is a median rather than a mean (so it is not dragged up by high earners), and it approximates what the filer personally forgoes.
• Use $44.50–$44.67 for aggregate claims — because that is what both Tax Foundation and NTU used, and citing the opposition's own valuation rate is stronger than picking our own. But say what it is: an employer cost figure including benefits, which is more than a worker takes home.
• Show $7.25 as the floor. "Even valued at the federal minimum wage — unchanged since 2009 — the average filer's lifetime is worth $5,655 to the tax-prep industry's continued existence" is a rhetorically useful floor precisely because nobody can argue it down.

4.1 Total lifetime cost: time + money

(Business "@ NTU/TF rate" uses Tax Foundation's business rate of $56.67: $81,605 + $37,200.)

Headline: about $36,500 of a life, per average filer, in constant dollars. Roughly half time, half cash — which is a useful thing to say, because it pre-empts "but the software is cheap now."

---

5. Household and both-spouses variants

5.1 The trap, first

IRS burden is measured per return, not per person. A married couple filing jointly files one return and incurs one 13-hour burden. Doubling the per-person figure for a married couple is wrong and it is the most likely arithmetic error in this whole document. Do not write "26 hours per couple."

5.2 A transparent household model

Two adults. Each files singly from age 20 until marriage at age M, then jointly from M to 80.

Household lifetime returns = 2 × (M − 20) + (80 − M) = M + 40

Arithmetic for the married-at-30 row: 2 × 10 + 50 = 70 returns; 70 × 13 = 910 hours; 910 ÷ 24 = 37.9 days; 70 × $290 = $20,300; 910 × $24.51 = $22,304.10.

5.3 Household opportunity cost, married at 30

The payment stream is $580/yr for 10 years (two returns) then $290/yr for 50 years.

• $290/yr for all 60 years: 290 × 353.584 = $102,539
• Extra $290/yr for years 1–10, compounded to year 60: 290 × [(1.05^10 − 1)/0.05] × 1.05^50 = 290 × 12.5779 × 11.4674 = $41,829
• Total = $144,368 in today's dollars

The early payments compound longest, which is why $20,300 of fees turns into $144,368 rather than the $102,539 that $17,400 of fees turns into.

5.4 What to say about a family

The clean, true, family-scale claim is:

A married couple who each filed alone through their twenties will spend 910 hours on federal tax returns — 38 days, around the clock — and $20,300 in fees. Invested instead, those fees are $144,000 by retirement.

Two never-married average filers are the only two-person case that reaches "two months" (65 days). Say "a couple," not "a person," if you use it.

---

6. The aggregate, for scale

6.1 Individual income tax filing — our scope

Source: NTU Foundation Feb 2026, Table 1 ("Individual Income Tax Compliance Burden, 2017–2025").

Derived scale comparisons (our arithmetic, from the government-derived hours):

• Full-time-equivalent jobs: 2.129 bn ÷ 2,080 hrs/yr = 1,023,558 FTE — over a million full-time jobs' worth of labor, producing nothing, every year.
• Person-years of round-the-clock time: 2.129 bn ÷ 8,766 hrs/yr (24 × 365.25) = 242,870 person-years. TY2025: 1.948 bn ÷ 8,766 = 222,222 person-years.
• Against incarceration: 1,852,900 people were held in US prisons and jails at yearend 2023 (Bureau of Justice Statistics). Individual tax filing consumes 242,870 ÷ 1,852,900 = 13% as much human time as the entire American carceral system.

The cleanest form of that last one, and the one that connects to the hero:

Every year, filing individual income tax returns costs Americans about 243,000 person-years — the same as taking 243,000 people and locking them up, around the clock, for a year. Every year.

Caveat that must ride along: this is the individual income tax figure (2.1 bn hours). Do not compute it from the 7.1 bn or 7.9 bn all-federal-compliance figures — those include corporate and payroll tax compliance and would inflate this to 810,000–901,000 person-years on a scope we are not arguing about.

6.2 All federal tax compliance — context only, label the scope

Tax Foundation notes $546 bn is "nearly 2 percent of GDP" and "26 times the IRS's 2024 annual budget."

Use these only when the sentence is about the tax code as a whole. The moment the sentence is about filing your 1040, drop to §6.1.

---

7. What automation can actually take back — the honest ceiling

The numbers above are what mandatory filing costs. They are not what pre-filling recovers. Publishing the first as though it were the second is the fastest way to lose the argument.

7.1 Ceiling from the return mix

Non-business filers are 71% of returns but only 44.9% of the hours:

0.71 × 8 = 5.68 weighted hours, out of 12.64 total → 5.68/12.64 = 44.9% (TY2024) 0.71 × 8 = 5.68, out of 11.77 → 48.3% (TY2025)

And only 39% of the dollars: 0.71 × 160 = $113.60 out of $293.40 = 38.7%.

So even perfect pre-population of every single non-business return removes at most about 45–48% of individual filing hours and 39% of the fees. The rest requires simplifying the tax code, not automating it.

Lifetime, for a person who stays non-business the whole way: at most 480 hours (20 days) and $9,600 recovered.

7.2 Ceiling from the accuracy research

Goodman, Lim, Sacerdote & Whitten (Treasury economists, NBER w30008 / NTJ 2023) find 42–48% of returns could be accurately pre-populated from current-year information returns plus the prior-year return, with accuracy falling as income rises and as dependents increase. That independently caps the reachable population at roughly the same place as §7.1's return mix — for different reasons, so it is corroboration, not double-counting.

7.3 The measured floor

The only number here the government actually measured rather than modeled:

$5,600,000 in preparation fees saved ÷ 140,803 Direct File pilot returns = $39.77/return

Over 60 filings: $2,386 in today's dollars. That is the floor — for a tool that was stage one (a free filer), not stage two (pre-filled), and only for the people who used it.

The honest bracket for lifetime out-of-pocket savings from a simple filer is therefore $2,400 (measured, stage one) to $9,600 (modeled ceiling, stage three).

7.4 How to say this without hedging into mush

Pre-filling does not delete the tax code. At best it takes back the simple half — about 45% of the hours Americans spend on 1040s, and 39% of the fees. That is still 20 days of a life and $9,600, per person, and it is money and time the government could hand back tomorrow for about $10 a user.

---

8. The numbers to lead with

Five claims, each one line, each with source and mandatory caveat. Advocacy estimates are flagged.

1. "The average American spends 13 hours and $290 a year filing a federal tax return." - Source: IRS taxpayer burden model, tax year 2024, compiled by NTU Foundation (Feb 2026). - Type: government model figure, relayed by a secondary source that opposes Direct File — which is why it is the right citation. - Mandatory caveat: it is a mean, not a median. Many filers pay $0; some pay $800. Say "the average filer," never "you." - Freshness caveat: the TY2025 update is 12 hours, $290. Both are fine; name the year.

2. "Over a filing life that is 780 hours — 32.5 days, counted the way a jail sentence is counted. More than a month." - Source: arithmetic on #1. 13 × 60 = 780; 780 ÷ 24 = 32.5. - Type: our derivation, from a government figure. - Mandatory caveat: state the conversion — 24-hour days — in the same breath. Never say "60 workdays, so basically three months."

3. "$17,400 in tax-prep fees over a lifetime — about $102,500 if you'd invested it instead." - Source: $290 × 60 = $17,400 (constant 2024 dollars); FV at 5% real over 60 years. - Type: our derivation; the 5% real return is our assumption and must be stated. - Mandatory caveat: $17,400 is today's dollars, not a nominal lifetime sum; $102,500 is also in today's dollars because the 5% is a real return.

4. "If you have business income — 29% of filers — it's 24 hours a year. That's 1,440 hours: exactly 60 days. Two months." - Source: IRS TY2024 per-return burden via NTU; 24 × 60 ÷ 24 = 60.0 days. - Type: government figure + arithmetic. - Mandatory caveat: say "tax year 2024." The TY2025 estimate is 21 hours, which is 52.5 days, and the claim no longer reaches two months.

5. "Every year, filing individual income tax returns costs Americans about 243,000 person-years — the equivalent of jailing 243,000 people, around the clock, for a year." - Source: NTU Foundation Feb 2026 Table 1 (2.129 bn hours, TY2024) ÷ 8,766 hours/year. - Type: our derivation from a government-derived aggregate. - Mandatory caveat: this is the individual income tax figure. Never compute it from 7.1 bn or 7.9 bn — those are all federal tax compliance, mostly business.

Bonus, and the one that closes the hero loop:

6. "The IRS's own estimate for running pre-filled filing at national scale is about $10 per person per year." - Source: IRS Report to Congress (May 2023) via CRS IF12654 — $249M/year at 25M users. $249,000,000 ÷ 25,000,000 = $9.96. The narrow case is $64M ÷ 5M = $12.80. - Type: government figure + arithmetic. - Mandatory caveat: critics (NTU) compute $54–$138 per return using pilot-scale costs, which is a fair criticism of a program kept small on purpose. Concede the per-unit point and make the fixed-cost argument.

Numbers to keep off the homepage

---

9. How a hostile reporter attacks this, and the answer

---

10. Corrections needed in the research doc

Three material errors found in ../../docs/research/policy-03-tax-filing.md, in descending order of danger:

1. Line 18: "NTU Foundation's 2025 estimate for income-tax compliance alone is 7.1 billion hours and $464 billion." The 7.1 bn / $464 bn figure is all federal tax compliance, not income tax alone. NTU's own individual income tax figure is 2.129 billion hours and $143.79 billion (TY2024, their Feb 2026 Table 1) — cross-confirmed by Tax Foundation's "individual income tax return hours: 2.249 billion" for the prior vintage. This is a 3.4× overstatement on the exact metric the site is about, and it is the error a hostile reporter is most likely to find. Fix before publishing anything.

2. Line 17: "Nonbusiness filers average about 8 hours and $140." Hours are right; the dollar figure is $160, in both TY2024 and TY2025 (NTU Feb 2026, Table 2).

3. Line 17: "filers with business income average about 25 hours and $530." Published figures are 24 hours / $620 (TY2024) and 21 hours / $610 (TY2025). Neither matches 25/$530. The hours understate and the dollars understate substantially.

Also worth adding to the research doc: NTU's and Tax Foundation's valuation rates ($44.67 and $44.50/hr) are near-identical, and NTU's 71/29 non-business/business return split — both are useful and neither is currently captured.

---

Appendix: constants used

Sources fetched and verified for this document

• NTU Foundation, "Individual Income Tax Compliance Burdens Ease for the New Filing Season," 2026-02-02 (Tables 1 and 2) — https://www.ntu.org/foundation/detail/individual-income-tax-compliance-burdens-ease-for-the-new-filing-season
• NTU Foundation, "Average American Spends 13 Hours and $290 to File Taxes" (7.1 bn hrs, $148 bn out-of-pocket, $464 bn total, $44.67/hr) — https://www.ntu.org/foundation/detail/taxpayers-will-spend-71-billion-hours-464-billion-on-tax-compliance-in-2025
• Tax Foundation, "Tax Compliance Costs the US Economy $546B Annually" (7.9 bn hrs, $133.3 bn, $546.1 bn, $44.50 / $56.67 rates, 2.249 bn individual hours) — https://taxfoundation.org/data/all/federal/irs-tax-compliance-costs/
• BLS, Occupational Employment and Wages, May 2025 (median hourly wage all occupations $24.51) — https://www.bls.gov/news.release/ocwage.t01.htm
• US DOL, Minimum Wage ($7.25 since 2009-07-24) — https://www.dol.gov/agencies/whd/minimum-wage
• BJS, Correctional Populations in the United States, 2023 (1,852,900 incarcerated) — https://bjs.ojp.gov/library/publications/correctional-populations-united-states-2023-statistical-tables

Not re-fetched this pass, taken from the research doc: IRS Direct File pilot closeout ($5.6M fees saved, 140,803 returns); CRS IF12654 ($64M–$249M mature cost); Goodman et al. NBER w30008 (42–48%); GAO-22-105236 (Free File <3%).