The TaxTools Network Standard

1. Official sources

Every rate, threshold, cap, band and cut-off used in a calculation is taken from the body that set it — the ATO, the IRS and the state revenue departments, the CRA and the provincial finance ministries, HMRC and GOV.UK, or Inland Revenue — or from the legislation, budget measure or public notice that changed it. Nothing is entered from memory, and no accountant's blog, news report or competing calculator is ever the basis for a figure.

Secondary sources have exactly one use: cross-checking. If an accounting firm's summary disagrees with our reading of the official page, that is a signal to re-read the official page, not a source to cite. The pages that carry a substantive number name the specific official documents behind it in a footer, so the claim can be checked rather than trusted.

2. Transparent calculation

A tax answer that arrives as a single number is not much use: the reader cannot tell whether it applies to them. Where a calculation has more than one step, the calculators show the steps — which bands were applied to which slice of income, which thresholds were crossed, which levies, offsets and surcharges were added, and what remains.

Where the official guidance is ambiguous, or depends on something we cannot observe (residency, an allowance transferred to a spouse, a marginal-relief edge case), the calculator states the assumption it made rather than hiding it. Assumptions are shown next to the result, not buried in a disclaimer.

3. Year-versioned rules

Tax rules are not a single set of numbers that gets edited each year; on these sites they are stored per tax year, in one place per site, and consumed from there by every calculator, article and page that mentions them. That is why a rate change lands in one file rather than being chased through prose.

Superseded years are kept, not overwritten. A calculator that offers a tax-year selector still computes last year's answer with last year's rules, because someone amending a past return needs last year's answer. Pages written about a particular year stay pinned to that year even after the year turns over.

4. Checked against official examples

Results are checked against the tax authorities' own published worked examples. Behind that, the network's calculation logic is separated from its interface specifically so it can be tested, and it is: 1,200+ automated test files across the five sites, run on every change before anything is published. Tests concentrate on the places tax code actually breaks — the exact income at which a band, threshold or phase-out begins, the dollar on either side of a cap, and the years that are no longer current.

Every site also carries a smoke test that fails the build when a tax-year rollover has been applied to some of a site's calculators but not all of them, because a half-applied rollover is the failure mode that is hardest to notice by eye.

5. Independence

The network is funded by advertising on member sites. The sites take no affiliate commission, no sponsorship, no payment for coverage and no payment for placement, and no advertiser has ever seen a calculator, an article or a result before it was published. There is no product being sold behind any answer, so no answer needs to point anywhere in particular.

Calculations run in the reader's own browser. Financial figures typed into a calculator are not transmitted to a server, and no result is stored or sold. This site — the network site you are reading — carries no advertising, no analytics and no third-party scripts of any kind.

How to check any of this

These are claims about behaviour, so they should be verifiable from the outside. The methodology page sets out the source hierarchy and the update process in detail; the accuracy page sets out what is tested and publishes the current counts; the corrections policy covers what happens when a number turns out to be wrong. If something on a member site does not match what is written here, that is a defect, and reporting it is the fastest way to get it fixed.