California’s 20-Year Tax Collection Clock Has a Constitution Problem

A 1932 amendment says every tax is conclusively presumed paid after 30 years. Ninety-plus years later, no court has decided whether that promise reaches your income taxes. A bill now advancing through the Legislature would fix part of the problem, while leaving the constitutional question wide open.

If you deal with California income tax, you know the rule on the books. The Franchise Tax Board has 20 years to collect a delinquent liability, and then it is supposed to be over.

That limit appears in Revenue & Taxation Code § 19255, added by Assembly Bill 911 (Chu), chaptered in 2005 and operative as of July 1, 2006. The clock starts when the latest liability for a tax year becomes “due and payable,” and once 20 years have elapsed, the FTB may no longer collect the debt, which is abated by operation of law. Before 2006, California had no general statute of limitations for collecting income tax. The State could chase an old bill indefinitely, which is precisely the open-ended exposure § 19255 was meant to close.

There is a catch, and it is why any of this matters. The FTB reads the statute to allow almost any later assessment (such as a small fee or a penalty) to restart the 20-year clock. Applied that way, the “limit” can run forever. (More on that below.)

So step back from the 20-year rule and ask where it came from. The Legislature didn’t choose that number in a vacuum. It anchored § 19255 to a much older and much broader promise, one written into the California Constitution, that may set a hard ceiling on the collection of any tax at all.

Buried in Article XIII of the California Constitution is a sentence that sounds almost too good to be true for any taxpayer carrying an old liability.

Every tax shall be conclusively presumed to have been paid after 30 years from the time it became a lien unless the property subject to the lien has been sold in the manner provided by the Legislature for the payment of the tax.

Article XIII, Section 30 opens with two words that ought to settle many arguments. Every tax. Read literally, the State’s power to collect any tax (income, sales, franchise, excise) expires after three decades, full stop.

Put in practical terms, the question is whether California can keep collecting an income-tax liability more than 30 years after it became a lien.

So here is the puzzle. In more than 90 years, no California appellate court has ever squarely held that this 30-year presumption applies to anything other than property taxes. It hasn’t held the opposite either. The entire reported history of the provision unfolds in disputes over real property, and its implementing statute sits quietly inside the property-tax division of the Revenue and Taxation Code. This promise has never actually been tested at its edges.

This piece lays out both readings as fairly as I can, then explains why I think the better one is also the most obvious one. Every tax means every tax. But the contrary argument is serious, and it has only gotten sharper. The question is also live in a way it has not been for years. A bill to curb the collection abuse that grows out of all this is further along than any of its predecessors, and even if it passes, it will not settle what the Constitution requires.

The text and the tension inside it

Section 30 contains two ideas pulling in opposite directions.

The first pulls broad. “Every tax” is categorical. The drafters knew how to limit a provision to real property. Neighboring statutes do exactly that, speaking of a tax “on real property,” and here they didn’t. When the People adopt unqualified language, courts ordinarily give it its ordinary, unrestricted meaning.

The second pulls narrow. The escape hatch, “unless the property subject to the lien has been sold in the manner provided by the Legislature,” is written in the grammar of secured real estate. It presupposes a specific property that the Legislature can sell to collect the tax. That describes the property-tax system perfectly. It does not describe income and sales taxes at all.

Everything in the debate flows from that tension.

Where Section 30 came from

Section 30 is a Depression artifact, and its history explains both its sweep and the fight over its reach. When real estate values collapsed in 1932–1933, property-tax delinquency in parts of the country reached 20–25 percent, starving local governments and school districts that depended on property taxes. Hard-pressed owners organized. Taxpayers’ leagues formed by the thousands, and in the most dramatic case, Chicago property owners staged an organized tax strike. (David T. Beito’s Taxpayers in Revolt: Tax Resistance During the Great Depression (Univ. of North Carolina Press, 1989) is the standard account.)

States responded with a wave of debtor relief, extending tax-sale deadlines, waiving penalties and interest, and allowing back taxes to be amortized to head off mass foreclosures. (See Wade S. Smith, Recent Legislative Indulgences to Delinquent Taxpayers, 3 Law & Contemp. Probs. 371 (1936).) And because the property-tax base had cratered, the same decade saw states permanently pivot to the revenue sources we now take for granted. The modern state income and sales taxes took firm root in the 1930s.

California’s voters were part of that wave, and Section 30 is one of its artifacts. Its immediate, practical target was real property. A 30-year absolute cutoff was meant to clear ancient, uncollected tax liens that were clouding land titles and locking parcels out of the market, forcing counties to enforce within a generation or forfeit the claim.

That origin story is genuinely double-edged, and both sides of the “every tax” debate reach for it. The narrow reading seizes on the title-clearing purpose. If the point was to unstick real estate, the provision was about property. The broad reading answers that the era’s finality impulse was never parcel-specific. It arrived hand in hand with the new income and sales taxes, and, as we’ll see, the 1932 ballot pamphlet told voters in plain terms that the measure applied to “every tax.” Hold both thoughts. They frame everything below.

What the statute does, and where it lives

The Legislature carried Section 30 into effect through Revenue & Taxation Code § 2195 (“Cessation of lien after 30 years”). Once 30 years have passed from the time a tax becomes a lien, if the lien hasn’t otherwise been removed, it ceases to exist, the tax is conclusively presumed paid, and the records officer marks it “Conclusively presumed paid.” (Amended Stats. 1997, ch. 546.)

Two features of § 2195 matter enormously, and both point toward property.

First, its address. Section 2195 sits in Division 1, Part 6, Chapter 2 of the Revenue and Taxation Code, the property-tax division, surrounded by provisions that are unmistakably about real property.

§ 2187 provides that a tax on real property is a lien against the property assessed.

§ 2192 (“Attachment of tax liens”) and § 2192.1 (“Priority of lien on real property”) fix when the lien attaches and where it ranks.

§ 2193 makes the lien operate as an execution duly levied against the property.

§ 2194 (“Satisfaction of judgment and removal of lien”) removes the lien only when the tax is paid or legally canceled, or the property is sold to satisfy the lien (and a sale of a tax certificate under § 4521 does not count as payment). (Amended Stats. 1995, ch. 189.)

§ 2196 covers the removal of an erroneously filed or invalid lien.

Second, its wiring. Section 2195 expressly does not apply to property for which a power to sell has been recorded for nonpayment. That carve-out cross-references the property-tax foreclosure machinery. Under § 3691 (“Power to sell”), once property has been tax-defaulted for five years (three for certain nonresidential commercial property), the tax collector “shall have the power to sell” it, and under § 3691.1 the collector records a Controller-prescribed notice to start that clock. (Both amended Stats. 2018, ch. 119.)

In other words, the 30-year presumption is hardwired into a sell-the-parcel enforcement system. That is the single strongest fact for the narrow reading, and an honest analysis has to put it front and center.

It is worth being precise about what this placement proves, though. It shows where the Legislature first implemented Section 30. It does not, by itself, amend the broader words the voters actually ratified, a distinction the rest of this piece turns on.

The case law and its near-misses

If you look for the case that decides whether Section 30 applies to income taxes, you won’t find it. What you find instead is a set of decisions that circle the question without answering it.

Paul v. Los Angeles County Flood Control District (1974) is the only decision to actually apply § 2195. A parcel’s tax became a lien in 1920, and the State waited until 1951 to deed the property to itself. Because the 30-year clock had run, the tax was conclusively presumed paid, and the deed was void. The court’s entire analysis is built around a real-property tax lien. It quotes § 2195’s “any tax becomes a lien” language but never asks whether “any tax” could mean a sales or income tax.

Home Owners’ Loan Corp. v. Hansen (1940) confirms that a non-property tax lien (there, a sales-tax lien) is purely a creature of statute, not of constitutional property-tax principle. No mention of Section 30.

Fullerton Oil Co. v. Johnson (1935) is quietly important. It confirms that franchise (and, by extension, income) taxes do “become a lien” under California law on a date fixed by the Legislature. That matters because Section 30’s trigger is the moment a tax “became a lien,” a trigger that, on its face, can fire for these taxes too. The court just never connected the dots to Section 30.

Agnew v. State Board of Equalization (1999) and Aronoff v. Franchise Tax Board (1963) run counter to the narrow reading. In both, the California Supreme Court applied other Article XIII provisions (§ 32 and its predecessor § 15) to sales and income taxes. So the Court plainly does not treat Article XIII as a property-taxes-only zone. It just hasn’t said so about § 30.

Mitchell v. Franchise Tax Board (1986) illustrates the structural difference that drives the whole debate. There, the FTB filed an income-tax lien to encumber “all property owned” by the taxpayers, as a general lien under § 19221 operates, rather than a single assessed parcel, as a property-tax lien does. (The court’s holding was narrower. The FTB was immune from the resulting damages suit under Government Code § 860.2.)

And for the property-tax sale machinery in action, see Lien Ly v. County of Alameda (2010). Note, though, that it is an unpublished opinion and, under California Rules of Court, rule 8.1115, generally may not be cited as precedent. It is illustrative only.

The narrow reading, at full strength

Put the strongest version on the table. The narrow reading isn’t a straw man. It rests on four real points.

The proviso speaks only property. “Unless the property… has been sold in the manner provided by the Legislature” assumes the existence of a specific asset that the State can foreclose and sell. For income or sales taxes, there is no such asset and no such sale. A reader can fairly conclude that the drafters were picturing real estate throughout.

The architecture is all property. As shown above, § 2195 sits among property-tax statutes and is directly linked to the § 3691 power-to-sell system. Non-property state taxes run on an entirely separate track. The state tax lien under Government Code §§ 7170–7172 lasts 10 years and can be renewed by re-recording, with no 30-year cessation rule anywhere in sight.

Other Article XIII exemptions are read as property-only. In Estate of Simpson (1954), the Supreme Court read the Article’s exemption and uniformity language as confined to property taxes, even though the text wasn’t explicit. Context, on this view, can quietly narrow a broadly worded clause.

The purpose was to clear land titles. As the history shows, Section 30’s practical target was ancient property-tax liens clouding real estate. A provision built to unstick land, the argument goes, is naturally read to reach only taxes that burden land.

Taken together, the narrow reading says that Section 30’s facially broad “every tax” was always meant to operate inside the property-tax lien-and-sale system, and nowhere else.

The broad reading, at full strength

Now, the other side, which I think is stronger.

The words are unqualified. “Every tax,” not “every property tax.” The drafters used the limiting language elsewhere and omitted it here. Courts don’t lightly read a missing word back into a constitution the People ratified.

The voters were told it meant everything. Section 30 was added to the Constitution as the 1932 Tax Liens Amendment (Proposition 16). The 1932 Voter Information Guide shows both sides describing a sweeping rule. The official argument against warned, in so many words, that “public utility taxes, corporation taxes, bank taxes, or any other kinds of taxes” left unpaid for 30 years would be canceled. (See also the measure’s ballot summary.) Opponents understood “every tax” to mean every tax, and under California law, the voters’ understanding informs the provision’s scope.

The Legislature has treated Section 30 as reaching income tax. When it enacted R&TC § 19255 in 2005 (via AB 911), the Legislature capped FTB collection of income taxes at 20 years as an expression of a finality principle rooted in Article XIII, Section 30. That was a policy of repose, with the constitutional tie more implied than declared. The 2026 record is sharper. The Senate Revenue and Taxation Committee’s analysis of AB 1519, a bill to amend income-tax collection, opens by describing Section 30 as “a statute of limitations for state tax liens.” That framing runs counter to the property-only reading in two ways. Property taxes are collected locally, so describing Section 30 in the language of state tax liens points toward the state-administered taxes the FTB collects, not the county property lien. And the property-only reading is one the FTB itself long held. Internal FTB records obtained through a Public Records Act request show the agency determining, as early as 2004, that Section 30 “applies only to liens on real property” and does not bar it from collecting an income-tax liability more than 30 years old. (FTB internal memorandum, Aug. 26, 2004, produced in FTB Public Records Act Response №23–00958 (Dec. 7, 2023).) A legislative committee’s move to place income-tax collection under Section 30’s umbrella is a meaningful step away from the property-only reading. It is not a holding, but it is the kind of co-equal-branch construction courts weigh under Lungren v. Deukmejian (1988).

Article XIII is not a property-only article. The Supreme Court has already applied its other provisions to income and sales taxes, so the section’s placement cannot, on its own, confine “every tax” to property.

The proviso is a dormant exception, not a hidden limit. The “unless… sold” clause lifts the presumption only when the secured property has been sold to collect the tax. It marks off one situation; it does not quietly shrink the class of taxes the rule reaches.

Weighing it, and why “every tax” should win

The narrow reading’s best move is the one I flagged at the start. Section 2195 is welded to the § 3691 sell-the-parcel system, and non-property taxes ride a completely separate Government Code regime with no cessation rule. If the machinery is all property, the argument goes, the constitutional rule must be too.

Practice is not text, and machinery is not reach.

I don’t think that follows for six reasons.

Machinery is not reach. The wiring between § 2195 and § 3691 describes how the Legislature implemented Section 30 for the one tax that is collected by selling a parcel. It is an implementation choice, not the outer boundary of a constitutional command that the voters wrote in categorical terms. Section 30 states a self-executing conclusive presumption, and § 2195 is one statute carrying it out, not its ceiling.

An exception cannot redefine the subject. The sharper version of the narrow reading says the whole sentence presupposes a saleable parcel, so income tax was never in view. The grammar says otherwise. “Every tax shall be conclusively presumed to have been paid after 30 years from the time it became a lien” is a complete command, and its subject is “every tax.” The clause “unless the property… has been sold” is a subordinate exception, and an exception withdraws the rule in a stated circumstance. It does not shrink the class the rule governs. Reading the sale clause to cut “every tax” down to property taxes lets the exception rewrite the subject, which inverts the ordinary relationship between a rule and its carve-out. The § 2195 carve-out only confirms the point. It mirrors Section 30’s own “unless… sold” proviso, suspending the presumption only when the sale path is live. Where a tax has no sale path, there is nothing to suspend, so the presumption stands.

The lien trigger genuinely fits income tax. Section 30’s operative trigger is the moment a tax “became a lien,” which is a moment of lien creation, and that moment has a determinate counterpart for income tax. California law is explicit that income and franchise taxes “become a lien” (the point of Fullerton), and the FTB’s lien under § 19221 arises on a fixed “due and payable” date. Notice where the property-specific language actually sits. The saleable “property subject to the lien” that the narrow reading leans on appears in the exception, “unless the property… has been sold,” not in the trigger. So the trigger maps onto the income-tax lien even though the exception is written in property terms. And this is not a resemblance the reader has to take on faith. When the Legislature enacted § 19255 to carry Section 30’s finality into the income-tax world, it ran the period from that same due-and-payable date, mapping Section 30’s lien-based rule onto the income-tax regime itself. To be candid, Section 30’s “lien” sits within property-sale grammar, and a skeptic can argue that the lien it contemplates is the parcel kind. But that objection lands on the exception, not on whether an income tax “became a lien” in the first place.

Separate collection statutes cut the other way. The Legislature set a 10-year limit for state tax liens in the Government Code, a 20-year limit for income taxes in § 19255, and, in 2016, authority for the FTB to discharge a liability “unpaid for more than 30 years” (Gov’t Code § 12437). None of that shows Section 30 ignores these taxes. It shows the Legislature legislating the details of the finality tax by tax and, in § 19255, doing so with Section 30’s finality principle in view. A skeptic will say the discharge statute would be pointless if Section 30 already extinguished the tax, but that provision is a discretionary write-off that lets the agency clear its books once a debt is uncollectible. A power to discharge is not the taxpayer’s constitutional right to repose, nor does it displace the constitutional floor beneath it. Fixed, non-renewable collection limits are also the national norm. Federal law caps collection at 10 years under 26 U.S.C. § 6502 with no administrative do-over, a policy the U.S. Supreme Court called “an almost indispensable element of fairness” in Rothensies v. Electric Storage Battery Co. (1946).

The structural argument proves too much. If placement and neighboring subject matter controlled, the Supreme Court could not have applied §§ 15 and 32 to income and sales taxes. But it did. The drafters’ deliberate choice of “every tax” is a more reliable guide than the company the section keeps in the code.

The purpose was property, but the words were general. The title-clearing history is the narrow reading’s other strong card, and it proves less than it seems. Clearing clouded titles was the pressing problem of 1932; it doesn’t follow that it was the limit of what the voters enacted. The decade’s drive for tax finality was general, not parcel-specific. It accompanied the very shift to income and sales taxes that defines 1930s public finance, and the ballot arguments described the rule as applying to utility, corporate, and bank taxes alike. A measure can be prompted by a property-tax crisis and still be written (and sold to voters) in categorical terms.

And to be fair to the other side, none of this makes the result certain. Every decided case involves property taxes; the implementing machinery is thoroughly property-specific, and a court inclined to follow that grain (reading the proviso as a window into a property-tax frame) could reasonably reach the narrow result, especially with no decision yet adopting the broad view. The honest description is that the text and history favor “every tax,” while the unbroken practice favors property. That gap is exactly what makes this an open question rather than a settled one.

The clock that never runs out

For decades, the scope question stayed theoretical, because the property-tax presumption runs automatically (the collector just marks the record) and income-tax fights get funneled into refund procedures. That is changing, and the reason is administrative.

The pressure point is § 19255’s 20-year cap on FTB income-tax collection. Public Records Act disclosures in 2023–2024 revealed that FTB systems treat almost any later assessment, even a small fee or penalty, as restarting the 20-year clock. Applied that way, a statute of limitations becomes a never-ending mechanism for collection. A taxpayer who owed tax in 2001 can find the State still pursuing it in 2041, kept alive by nothing more than a re-recorded lien fee.

The long road to a fix

None of this is new to the Legislature, which has spent two decades trying to make the 20-year limit mean what it says.

The story starts with the limit itself. AB 911 (Chu) created the 20-year cap in 2005, operative the following July, replacing an era when the FTB could pursue an old income-tax bill with no deadline at all. But the statute defined “tax liability” to include interest, penalties, and fees, and the FTB read that definition as allowing each new charge to restart the clock. Within a decade, practitioners realized the limit was, in practice, no limit.

The first serious repair was AB 357 (Nazarian, 2019). It would have excluded penalties and fees from the definition, and it passed both houses without a single dissenting vote. Then Governor Newsom vetoed it. His objection was purely fiscal. The bill, he wrote, would “significantly” limit the FTB’s “ability to collect valid tax liabilities” and do so “at a significant cost to the state general fund.” He said nothing about the Constitution. That is a real budgeting concern, but it carries little weight if Section 30 already bars collection after the constitutional period, because revenue the Constitution places off-limits was never the state’s to keep.

A follow-up, AB 2369 (Nazarian, 2020), tried again and died in committee as the COVID session collapsed. The idea then sat until this session, when it was revived through a gut-and-amend. AB 1519 (Gipson), a bill that began as an unrelated tax-filing-notice measure, was rewritten in the Senate on April 28, 2026, into the collection fix.

This time, the fix is closer than it has ever been. AB 1519 redefines “tax liability” to exclude interest, penalties, costs, and fees and requires those charges to lapse together with the underlying tax. It passed the Assembly, cleared the Senate Revenue and Taxation Committee, and now sits in Senate Appropriations with a hearing set for August 3, 2026. The fiscal objection that sank AB 357 also looks smaller now. Using the FTB’s own figures, the committee's analysis estimates that the fix would discharge only about $100,000 in its first year and minimal amounts thereafter. It drew support from the California Chamber of Commerce, the California Taxpayers Association, CalCPA, and the Taxation Section of the California Lawyers Association, with no opposition on file. And the same analysis, reviewing an income-tax bill, opens by describing Article XIII, Section 30 as the constitutional statute of limitations that stands behind the entire scheme.

Why a statute won’t finish the job

Suppose AB 1519 passes. The reset abuse would stop, and penalties and fees could no longer keep the clock running. But the deeper problem would remain because a statute can be narrowed and has gaps.

Two gaps survive AB 1519. The bill still lets a later tax assessment for the same year push the 20-year clock forward, and it keeps the tolling provisions that can suspend the clock for years at a time. Either path can still carry collection past the 30-year mark. And AB 1519 is only a statute, so a future Legislature can amend it, and the FTB will keep interpreting it. Only Article XIII, Section 30 supplies an outer limit that no assessment, fee, tolling event, or later statute can exceed. That is why the constitutional question does not disappear even if the bill becomes law. If anything, the 20-year fight is the clearest evidence yet that the political branches cannot be relied on to guarantee finality on their own.

The bottom line

The better reading of Article XIII, Section 30 is the plain one. “Every tax” means every tax. The 30-year conclusive presumption of payment should apply to income, sales, franchise, and excise taxes, not just to liens on real property. That conclusion rests on the words the voters ratified, the way both sides described the measure in 1932, and the Legislature’s own reliance on Section 30’s finality, both when it capped income-tax collection in 2006 and again as it debates AB 1519 today.

What stands in the way is not a contrary holding (there isn’t one) but 90 years of property-tax-only practice and an implementing machinery built entirely around selling parcels. Those facts deserve respect, and I’ve tried to give them their due. But practice is not text, and machinery is not reach. Until a court is finally asked the question, the biggest promise in California tax law remains exactly that, a promise, conclusively presumed, and still waiting to be tested.

What happens next

The next milestone is not a courtroom but a committee room. AB 1519 faces the Senate Appropriations Committee on August 3, 2026, and the weeks leading up to that date are when its fate will be effectively decided. Anyone who advises California taxpayers on aging liabilities has a stake in the outcome. If you have clients caught by the reset, that is the window to make the experience known to the Legislature.

And whatever the committee does, the larger question stands. Whether “every tax” in the Constitution means what it says will not be answered by a statute. It is still open and still waiting for the right case.

This article is for general informational purposes only and does not constitute legal, tax, or other professional advice. Reading this article or contacting Galek Tax Law through this website does not create an attorney-client relationship. You should not act or refrain from acting based on this article without seeking advice from counsel regarding your specific facts.