Completion of the Filing Process · Accuracy
Miscalculations and recognition of duplicate entries
tax year · reviewed 2026-08-19 · I. Ohu
The rule
A miscalculation is not a small version of a bad position; it is a different kind of problem with a different procedure attached, and a practitioner who does not know the difference will give the wrong advice about a notice.
On the IRS’s side. Where the IRS notifies a taxpayer that, on account of a mathematical or clerical error appearing on the return, an amount of tax in excess of that shown is due and has been or will be assessed, that notice is expressly not a notice of deficiency (IRC § 6213(b)(1)). The taxpayer has no right to petition the Tax Court on it, the ordinary bar on assessment does not apply, and the IRS may assess and collect. The notice must set forth the error alleged and an explanation of it. The taxpayer’s remedy is a request for abatement within the statutory window under IRC § 6213(b)(2), which forces the Service back into ordinary deficiency procedures if it wishes to persist — with the notice of deficiency and the Tax Court petition right that come with them.
Section 6213(g)(2) defines what counts. It reaches more than arithmetic: an error in addition, subtraction, multiplication or division shown on the return; an incorrect use of an IRS table where the incorrectness is apparent from other information on the return; an entry inconsistent with another entry on the same or another schedule; an omission of information required to substantiate an entry; a deduction or credit exceeding a statutory limit expressed as a specified monetary amount, percentage, ratio or fraction where the items entering the limit appear on the return; and — the categories that matter most in practice — omitted or incorrect taxpayer identification numbers where the Code requires them.
On the preparer’s side. Reg. § 1.6694-2(e) supplies the framework for whether an error attracts a penalty. Its factors are set out in the figures table, and three of them speak directly to miscalculation: the nature of the error, the frequency of errors, and the materiality of errors. An isolated error — the regulation’s own example is “an inadvertent mathematical or clerical error” — generally qualifies for the reasonable cause and good faith exception. It does not qualify where the isolated error is so obvious, flagrant, or material that it should have been discovered during a review of the return, and the exception falls away entirely where there is a pattern of errors on a return even though any one error in isolation would have qualified.
Current figures
| Item | Content |
|---|---|
| Reasonable cause and good faith factors | 6 — the nature of the error causing the understatement, the frequency of errors, the materiality of errors, the preparer's normal office practice, reliance on the advice of others, and reliance on generally accepted administrative or industry practiceTY2026 |
| What a normal office practice must be | a system for promoting accuracy and consistency that, for a signing preparer, generally includes checklists, methods for obtaining necessary information from the taxpayer, a review of the prior year's return, and review proceduresTY2026 |
| New taxpayer declaration required if data changes by more than | $50 to “Total income” or “AGI”TY2026 |
| Or by more than | $14 to “Total tax,” “Federal income tax withheld,” “Refund” or “Amount you owe”TY2026 |
How it works in practice
Duplicates have three recurring sources, and each has its own check.
- The same document entered twice. A Form 1099 arrives by post and again in a client portal upload, or a spouse and a preparer each enter a brokerage statement. The check is a reconciliation against a document count, not a scan for unfamiliar entries — a duplicate looks entirely familiar.
- The same item claimed in two places. A vehicle expense taken both on Schedule C and as an unreimbursed employee cost; a payroll deduction claimed on the business return and again personally; a state tax payment counted both as an itemised deduction and as a business expense. Publication 1345 flags the underlying discipline for EROs reviewing wage statements: identify payroll deductions to ensure the expense is necessary and not claimed again on the statement, and identify deductions to ensure an expense is claimed once, for personal or business, not both.
- The same person counted twice. A dependent claimed on two returns, most often after a separation. This one the IRS catches by identification number, and it is squarely within the math error categories in § 6213(g)(2).
Rolling forward last year’s file is the most productive duplicate generator there is. A carried schedule that already contains an item, plus this year’s document entered fresh, produces a duplicate in a file that otherwise looks correct. Reg. § 1.6694-2(e)(4) names a review of the prior year’s return as an element of a normal office practice — its function is not only to catch what is missing but to catch what was carried.
Test the output against expectation. The reviews that catch computational errors are comparative rather than arithmetic: does the effective rate fit this income, does the refund track last year adjusted for what actually changed, does a phase-out that should have bitten appear to have bitten, does a total tie to the underlying schedules. A preparer who only checks that entries match source documents will confirm a return that is internally consistent and substantively wrong.
Correcting after the client has signed has a threshold. Where a return has been rejected and the correction changes the data by more than the amounts in the figures table, the taxpayer must sign a new declaration before retransmission. Inside both thresholds, the correction can go out on the existing signature. This is where a discovered miscalculation most often collides with process, and the temptation to fix and resend without going back to the client is exactly what the thresholds exist to control.
Advise on a math error notice correctly. The client’s instinct is that a notice means an audit and the Tax Court. It does not. There is no petition right on a math error notice, the assessment is not blocked, and the route is the abatement request within the statutory window — which, if made, puts the Service to deficiency procedures. Missing that window converts a correctable notice into an assessed balance.
The brokerage statement that arrived twice
Anneke Lindqvist-Baptiste uploads her documents to her preparer’s portal in February and posts the paper originals in March, worried the upload failed. Two people at the firm process the two batches. The return reports her brokerage income twice.
The error is invisible on review of the entries — each one matches a real document. What catches it is a count: the client has one brokerage account and the return carries two of its statements. The office practice element that would have prevented it is the “methods for obtaining necessary information from the taxpayer” that Reg. § 1.6694-2(e)(4) names — a single intake channel with a document log, rather than two people working two batches. The error overstates income, so it costs the client money rather than exposing her, which is why nobody would have found it in an examination either.
The vehicle claimed twice
Emmerich Sørensen-Achebe runs a consultancy and also has employment income. His vehicle costs are entered on Schedule C from his mileage log and again, by a different staff member working from his employer’s reimbursement summary, as a separate cost.
This is the “claimed once, for personal or business, not both” problem Publication 1345 directs EROs to look for when reviewing wage statements and payroll deductions. On the preparer penalty side it is not obviously an isolated inadvertent error: it is a duplicate deduction that a review comparing the Schedule C to the employment records would have surfaced, which brings it close to the “should have been discovered during a review” exclusion. The fix is structural — one person owns the expense schedule for a client with both income types, and a cross-check against the other schedule is a standing item.
The math error notice the client wanted to appeal
Perpetua Vandermeer receives a notice stating that an entry on her return was inconsistent with another entry and that additional tax has been assessed. She instructs her enrolled agent to appeal to the Tax Court.
That route does not exist on this notice. A math or clerical error notice is expressly not a notice of deficiency, there is no Tax Court petition right on it, and the assessment is not barred. What she can do is request abatement within the statutory window, which obliges the Service to proceed by deficiency procedures if it maintains the position — producing a real notice of deficiency, and with it the petition right she thought she already had. The advice that matters is about the window, and it is short enough that the conversation has to happen the week the notice arrives.
How this has changed
The math error authority has expanded steadily and quietly. It began as a narrow power to correct arithmetic and now reaches, under § 6213(g)(2), inconsistent entries, omitted substantiating information, limits computable from the face of the return, and — the category added and broadened repeatedly since the 1990s — omitted or incorrect taxpayer identification numbers where the Code requires one for a credit or a dependent. Each expansion moved a class of case out of deficiency procedure and into a notice with no petition right, which is why understanding the abatement request matters more now than it did.
On the preparer side, the six reasonable cause factors took their current form in the 2008 preparer penalty regulations, which also supplied the description of a normal office practice that this topic leans on. The description has aged well precisely because it is about process rather than technology: checklists, methods for obtaining information from the taxpayer, review of the prior year’s return, and review procedures describe what a firm does, not what its software does.
Exam focus
Know that a mathematical or clerical error notice is not a notice of deficiency, carries no Tax Court petition right, and does not bar assessment, and that the taxpayer’s remedy is a request for abatement within the statutory window which puts the Service to deficiency procedures. Know that the statutory definition reaches more than arithmetic — inconsistent entries, omitted substantiating information, limits computable from the return, and identification number problems. On the preparer side, know that an isolated inadvertent mathematical or clerical error generally qualifies for the reasonable cause exception, but not where it should have been discovered on review, and not where there is a pattern.
Check yourself
1. A taxpayer receives a notice assessing additional tax because of a clerical error on the return. What may the taxpayer do?
A. Petition the Tax Court within 90 days B. Request abatement within the statutory window, which puts the Service to deficiency procedures C. Nothing; the assessment is final on issuance D. Request an Appeals conference as of right before assessment
Answer: B. The notice is expressly not a notice of deficiency, so no petition right attaches to it.
2. Which is not within the statutory definition of a mathematical or clerical error?
A. An entry inconsistent with another entry on the same schedule B. An omission of information required to substantiate an entry C. A disputed valuation of donated property D. A deduction exceeding a statutory limit expressed as a specified monetary amount, where the items entering the limit appear on the return
Answer: C. A valuation dispute is a substantive disagreement requiring deficiency procedures.
3. A preparer makes one inadvertent arithmetic error that is obvious on the face of the return. Does the reasonable cause and good faith exception apply?
A. Yes; isolated errors always qualify B. Yes, provided the understatement is small C. No, if the error is so obvious, flagrant or material that it should have been discovered during a review D. No; the exception never applies to arithmetic
Answer: C.
4. What does the regulation describe as a normal office practice?
A. Owning current tax preparation software B. A system for promoting accuracy and consistency, generally including checklists, methods for obtaining information from the taxpayer, review of the prior year’s return, and review procedures C. Employing at least one credentialed practitioner D. Retaining returns for three years
Answer: B — and the factor helps only where the practice was actually followed on the return in question.
5. A duplicate entry causes a client’s income to be overstated. What is the practical consequence?
A. The IRS will identify and correct it through a math error notice B. The error is immaterial because no tax was understated C. The client overpays, and only the firm’s own review will find it D. The reasonable cause exception makes it irrelevant
Answer: C. Examination looks for understatements; nobody outside the firm is looking for this one.
Change log
- Initial draft.