Nine jurisdictions deleted a rule you may have already crossed
A sales tax threshold is not a setting. It is a dated law, and states change it.
Since January 2023, nine of the jurisdictions we monitor have removed their 200-transaction test and left only a dollar figure. If your orders met one of those tests before it was removed, you met it. The repeal did not reach backwards and un-happen the day you crossed.
This is the follow-up we promised in the Kentucky piece, and the principle is bigger than Kentucky.
The nine, and the date each one changed
Every one of these removed a 200-transaction test and kept a $100,000 dollar test. The date in the middle column is the boundary: sales before it are measured one way, sales after it another.
| Jurisdiction | Transaction test removed | Rule before | Rule from that date |
|---|---|---|---|
| South Dakota | 2023-07-01 | $100,000 or 200 transactions | $100,000 |
| Louisiana | 2023-08-01 | $100,000 or 200 transactions | $100,000 |
| Indiana | 2024-01-01 | $100,000 or 200 transactions | $100,000 |
| North Carolina | 2024-07-01 | $100,000 or 200 transactions | $100,000 |
| Wyoming | 2024-07-01 | $100,000 or 200 transactions | $100,000 |
| Alaska | 2025-01-01 | $100,000 or 200 transactions | $100,000 |
| Utah | 2025-07-01 | $100,000 or 200 transactions | $100,000 |
| Illinois | 2026-01-01 | $100,000 or 200 transactions | $100,000 |
| Kentucky | 2026-08-01 | $100,000 or 200 transactions | $100,000 |
Nine changes in about three and a half years, and the cadence has not slowed. This list is what our rule table actually holds today, back to a history floor of 2023-01-01. It is not the complete history of US economic nexus, which runs back to 2018, and we would rather state our floor than imply a depth we do not have.
Where the "200" comes from, because you should not take it on our word. Our source states the 200-transaction figure outright for Kentucky. For the other eight it records that the test was removed on the date shown, without restating the number that was removed. We carry the 200 forward from the current chart entry, and we carry the dollar threshold, the window and the marketplace treatment forward the same way. That is an inference, it is a reasonable one, and it is not the same kind of fact as Kentucky's. Every one of those eight rules says so on its own screen in the app, in the citation attached to the verdict, and it would be strange to disclose that in the product and not here.
Alaska belongs on the list with two caveats attached. Alaska has no statewide sales tax: its threshold is administered by municipalities through the Alaska Remote Seller Sales Tax Commission, so crossing it is an obligation to those municipalities rather than to the state, which is a different registration conversation. And we model the ARSSTC aggregate only. Alaska municipalities outside that program are not modelled here, the same way we do not model Colorado's home-rule cities or Louisiana's parishes.
Why this is the hard case, in one sentence
The transaction test is the one you cross without noticing, because it has nothing to do with how much money you made.
$100,000 is a number a seller feels. Two hundred orders of a $28 phone case is $5,600, which is a rounding error in a business doing $40,000 a month, and it used to be enough to meet the test in nine of these places.
So the population most exposed to a historical crossing is exactly the population least likely to go looking for one.
The worked example: same orders, two states, opposite answers
A seller ships a $28 phone case. During calendar 2023 they shipped these into two states, split across Shopify and Amazon:
| Orders | Amount | |
|---|---|---|
| Shopify | 80 | $2,240 |
| Amazon | 150 | $4,200 |
| Total shipped, both channels | 230 | $6,440 |
Identical volume into North Carolina and into Indiana. Now measure each one against the rule that was in force in 2023, which for both was "$100,000 or 200 transactions".
Both 2023 rules here are reconstructed rather than sourced directly, in the way described above, so read this as the best reconstruction our sources support for 2023 and not as a quotation from a 2023 statute. The shape of the lesson does not depend on the exact figures; the difference between the two states does, and that difference is stated identically on both the 2023 and the current chart entry.
North Carolina, calendar 2023. North Carolina counts marketplace sales toward your own threshold.
| Orders | Amount | |
|---|---|---|
| Total into North Carolina | 230 | $6,440 |
| Counts toward North Carolina's threshold | 230 | $6,440 |
| North Carolina's 2023 test | 200 transactions or $100,000 | |
| Result | Met, on the day the 200th order shipped |
Indiana, calendar 2023. Indiana does not count marketplace sales toward your own threshold.
| Orders | Amount | |
|---|---|---|
| Total into Indiana | 230 | $6,440 |
| Counts toward Indiana's threshold | 80 | $2,240 |
| Indiana's 2023 test | 200 transactions or $100,000 | |
| Result | Not met. 80 orders against a 200 test |
Same seller, same product, same 230 orders, opposite verdicts. Two things did the work and neither is visible in a Shopify report: which state is asking, and which rule was law at the time.
This is why we render two numbers for every state and never one. "230 orders into the state" is true in both places and decides neither. "80 orders count" is true in Indiana and wrong in North Carolina. There is no single number that is correct in both columns, so we do not invent one.
Now run the same seller through a tool that only knows today's rule
Today, in 2026, North Carolina's test is $100,000 and nothing else. Our seller is still running at roughly the same rate, about $6,400 a year into North Carolina.
A lookup against the current rule returns no nexus, comfortably, with a wide margin, and it is wrong about 2023. The 200-transaction test that seller met is not on North Carolina's page any more. It was law when they met it.
The failure is silent. There is no error, no warning, no gap in the output. The tool answers a question that was never asked ("do you meet today's test") in place of the one that matters ("did you ever meet a test"), and the two answers look identical.
An engine that stores one current row per state cannot do better than this, because the information needed to be right was overwritten the day the rule changed.
What we do instead
Every rule in StateWise is a row with an effective-from and an effective-to date. A 2023 period is evaluated against the 2023 row. A 2026 period is evaluated against the 2026 row. The same order history produces a different verdict in a different year, correctly, because the law was different.
When we find a crossing, we show you the date and the version of the rule that produced it. The sense of it is:
You met the North Carolina threshold on the day your 200th order was placed, on the 200-transaction test, which North Carolina changed on 2024-07-01.
The date says when. The rule version says under what. Neither is recoverable from a current-rules lookup, and together they are what an accountant needs to do anything useful.
One precision, since this article is about being precise: we date a crossing by the day the order was placed, not the day it shipped. "Shipped into the state" above is shorthand for "destined for that state", which is what the state's rule is actually about.
Every rule also carries the date we last verified it against its source, shown on the verdict. A rule we have not rechecked in six months says so on its own screen.
The four honest limits
One. We do not tell you the obligation is still live. How long a nexus obligation persists after you fall below a threshold is called trailing nexus, and it varies by state, and it is out of scope for us on purpose. We tell you precisely what we can establish from your orders and the statute: that you met the test, on this date, under this version of the rule. What follows from that is a conversation with your accountant or the state, and we would rather hand you a date and a citation than a guess.
Two. Our history floor is 2023-01-01. Rules before that date are not in our table, and we say so on screen rather than returning a confident "no crossing" that really means "we did not look". If you were selling at volume in 2019, a crossing may exist that we cannot see.
Three. We can only measure the orders you give us. Shopify we read directly. Amazon, Etsy, eBay and Walmart come in by CSV, and a channel you have not imported is a channel we count as zero. In the North Carolina example above, the Amazon half is 150 of the 230 orders. Leave it out and the verdict flips to "not met", which is the same wrong answer from the other direction.
Four, and this one bites the example above. What we read from Shopify directly reaches back 24 months, or 60 days if the historical-orders permission has not been granted. That window moves forward every day, so on the day you read this it may not reach 2023 at all. Five of the nine removal dates in the table (South Dakota, Louisiana, Indiana, North Carolina and Wyoming) are already older than a 24-month window, which means a crossing under those five is not something we can find from your Shopify data alone.
We are saying this plainly because the worked example above is exactly that case: 80 Shopify orders in calendar 2023, which is beyond what we can pull from Shopify today.
The way through it is the same CSV path the marketplaces use. Your own Shopify order export goes back as far as your store does, and importing it pushes our start date back to the earliest order in the file. So 2023 is reachable, it is just reachable by upload rather than automatically. The app shows you the exact date its coverage starts on, on screen, so you never have to infer it from this page.
What to check this week
- Find your 2023, 2024 and 2025 order counts by state, across every channel, not just the one with the nicest dashboard. Order counts, not just dollars. That is the test people miss.
- Compare each year against the nine dates above. A year that ended before a removal date is measured with the transaction test live. Two wrinkles. Illinois does not measure calendar years at all, it measures a trailing twelve months, so its dates do not line up with the other eight and a year-by-year count is the wrong unit there. And where a removal date falls inside a year (South Dakota and Louisiana in 2023, Utah in 2025), that year is split across two rules and neither one governs the whole of it.
- Check whether the state counts your marketplace sales. Twenty-eight of the forty-eight jurisdictions we monitor do. It is the difference between 80 orders and 230.
- If you find a crossing, get the date, not just the fact. A registration conversation goes differently when you can say which day and which statute.
Sources
- Rule versions, effective dates, thresholds, windows, sales basis and marketplace treatment for all 48 monitored jurisdictions: Sales Tax Institute, Economic Nexus State Guide (chart dated 2026-05-04), verified 2026-07-31 and re-verified against a second source 2026-08-03.
https://www.salestaxinstitute.com/resources/economic-nexus-state-guide - Kentucky HB 757, 2026 Regular Session, amending KRS 139.340. Enacted 2026-04-14 by veto override; nexus provision effective 2026-08-01.
https://apps.legislature.ky.gov/record/26rs/hb757.html - Alaska's statewide position and the municipal administration of its threshold: Alaska Remote Seller Sales Tax Commission.
https://arsstc.org/ - Louisiana is recorded by us as disputed on marketplace treatment, and both readings are rendered with their own number. Its 2023-08-01 transaction-test removal is not in dispute and is what places it on the list above.
The dates and thresholds here are the ones in our rule table on the date at the top of this page. We show you the arithmetic and the rule behind it so you can have an informed conversation with someone qualified to give you advice. The not-tax-advice notice below is the full version and it is rendered on every page here, so no article can be published without it.