How We Calculate Everything

1. 5-year LLC cost formula

For each US state, we compute the all-in 5-year cost of maintaining a single-member LLC as:

5-year cost = filingFee + 5 × (annualReportFee + raCostBaseline + franchiseTaxMin)

• filingFee — one-time Articles of Organization fee from the state Secretary of State.
• annualReportFee — the annual (or biennial / decennial, prorated) report fee. 0 for states without annual reports (Missouri, New Mexico, Ohio, South Carolina).
• raCostBaseline — registered-agent baseline of $50/yr. Reflects the cheapest realistic outside RA (Northwest Registered Agent budget tier). Some users pay $0 (own address) or up to $300 (premium services).
• franchiseTaxMin — the unavoidable annual franchise/privilege/business-tax minimum for an LLC in good standing. 0 in most states; $800 in California, $300 in Delaware/Maryland, etc.

What this does NOT include: publication cost (NY/AZ/GA/NE — see each state page), professional services (CPA fees), business licenses, sales-tax registration, foreign-LLC qualification fees, expedited filing surcharges. These are itemized on each state page individually.

Source code: function fiveYearCost in src/lib/tax.ts.

2. Federal income tax

We use the 2026 federal individual income tax brackets, which we estimate from IRS inflation indexing trend extending Rev. Proc. 2024-40 (which set 2025 figures). The 2026 figures will be officially published by the IRS in late October 2025; our values are best-estimate projections until then.

Standard deduction (2026 est.): $15,000 single / $30,000 married joint / $22,500 HoH.

Source: IRS Rev. Proc. 2024-40 (2025 inflation indexing).

3. State income tax

For pass-through LLC owners, we apply each state’s top marginal personal income tax rateto net business income. For C-Corp election we apply the corporate rate. The rates and structures are sourced directly from each state’s Department of Revenue and Tax Foundation cross-checks.

Important simplification: we model the top marginal rate, not the full progressive schedule. For low-income LLCs in heavily-progressive states (California, Hawaii, New York), this slightly overestimates the actual liability. The simplification is intentional: most users are above the median tax bracket, and the conservative estimate prevents pleasant-surprise undercounting.

4. Self-employment tax

SE tax = 0.9235 × netEarnings × (0.124 SS up to $176,100 + 0.029 Medicare + 0.009 additionalMedicare above threshold)

• Social Security wage base for 2026: $176,100 (estimated).
• Additional 0.9% Medicare applies above $200k single / $250k joint / $125k MFS.
• Half of SE tax is deductible above the line for federal income tax.

Source: IRS — Self-Employment Tax.

5. QBI deduction (Section 199A)

Section 199A allows pass-through LLC owners to deduct up to 20% of Qualified Business Income, subject to taxable-income thresholds and SSTB phase-outs.

2026 estimated thresholds:

• Full deduction: taxable income ≤ $197,300 single / $394,600 MFJ.
• Phase-out range: $50,000 single / $100,000 MFJ.
• Above phase-out: SSTB businesses (health, law, accounting, consulting, financial services, performing arts) get $0 QBI; non-SSTB businesses are subject to W-2 wage + UBIA limits.

State conformity to QBI varies. California, New Jersey, Pennsylvania, and Tennessee do not allow the federal QBI deduction to reduce state taxable income. New York applies its own state-level rules. Most other states default to federal AGI and inherit the QBI deduction automatically.

Source: IRS — QBI Deduction and 26 USC §199A.

6. S-Corp election breakeven

The S-Corp election (Form 2553) converts the LLC’s tax treatment so the owner-employee pays FICA on a reasonable salary, with the remaining profit passing through free of SE tax. We compute the breakeven point where SE-tax savings exceed payroll/admin overhead:

breakeven = reasonableSalary + 1500 / 0.1413

The 0.1413 factor is the effective SE-tax rate (15.3% × 92.35% Schedule SE adjustment) on the distribution portion that the election avoids. The $1,500 represents the floor for payroll-service + extra Form 1120-S + W-2 admin overhead. Real-world overhead is $1,500–$3,000.

7. Privacy & banking scores

The 1–10 scores on state and country data are heuristic, not statutory:

• Privacy reflects whether members/managers appear on public state filings. Wyoming, New Mexico, Nevada, and South Dakota score 8–9 (most anonymous). California, New York, Florida score 3–4 (members on public record). Score does NOT incorporate the FinCEN BOI report, which is non-public but mandatory in all states.
• Banking accessibility reflects how easily a Mercury/Relay/Wise Business account opens for a non-resident-owned LLC in that jurisdiction. Calibrated from user reports and major bank acceptance lists. Updated semi-annually.

8. Country data

For 30+ countries we track filing fee, annual maintenance, corporate tax rate, VAT/GST rate, minimum paid-up capital, banking accessibility, Stripe/PayPal/Wise availability, UBO publicity, and remote-formation eligibility. Each country page shows the official registry/authority link we used as the primary source, plus a Last verified date.

9. What we deliberately don’t model

• Local city / municipal business taxes (e.g., San Francisco GRT, NYC UBT) — only state-level taxes are in the dataset.
• FICA cap for high-earner S-Corp owners in years they have multiple jobs — assumed sole-source income.
• Foreign tax credits for non-resident scenarios — handled in the dedicated non-resident calculator, not the standard tax calculator.
• State estate, property, sales, and excise taxes — out of scope for income/formation modeling.
• Self-employed health insurance deduction — adds ~5–10% to take-home for some scenarios but requires assumptions about insurance premium amount.

10. Verification and audit

Every figure can be traced to a primary source linked from the relevant page. To verify any number:

1. Locate the state or country page (e.g., /state/wyoming).
2. Scroll to the “Source” or “Authoritative source” panel.
3. Click through to the government registry.
4. Compare the figure on that page to the figure in our dataset.

Discrepancies should be reported to corrections@fevantech.com per our editorial policy.

Read next

• Editorial Policy
• Primary sources we cite
• Full disclaimer