Identifier: H.B. 437, Utah Legislature, 2025 General Session. Enrolled from the 3rd Substitute as amended by Senate Floor Amendment 1.
Full title: Interdicted Person Amendments
Passed: March 7, 2025, the final day of the session. House 70–0–5 (Feb. 20); Senate 26–0–3 (Mar. 7); House concurrence 65–0–10 (Mar. 7).
Signed: March 27, 2025.
Effective: January 1, 2026.
Sponsors: Rep. Steve Eliason (chief sponsor); Sen. Jerry W. Stevenson (Senate sponsor).
Authority: State legislation. Amends Titles 32B (Alcoholic Beverage Control Act), 41 (Motor Vehicles), 53 (driver licenses and identification cards) and 76 (Criminal Code); enacts §53-3-236.
Track: Legislative
Plain-language summary: H.B. 437 lets Utah courts designate people convicted of driving under the influence as “interdicted persons,” barred from buying alcohol. The designation is mandatory for an extreme DUI and for the offenses of negligently injuring or killing someone while driving impaired. A designated person must surrender their driver license or ID card. Their replacement card carries the words “No Alcohol Sale” and a red stripe, and anyone may apply for such a card voluntarily. To make the marking matter, the bill requires that anyone selling alcohol, anywhere in the state, examine the buyer’s ID before every sale to check for interdiction. That includes IDs from other states and countries, which cannot carry the mark. On the final day of the session, a one-line Senate floor amendment struck the words “who appears to be 35 years of age or younger” from Utah’s proof-of-age statute. That made electronic ID scanning mandatory for every patron entering a bar or tavern, and for every purchase in a dining club or in a restaurant’s dispensing area. Everything takes effect January 1, 2026.
Verdict: H.B. 437 sets out to catch a few dozen court-designated people a year buying alcohol, which is already a crime for them, by checking the ID of every buyer in the state, yet the check can only detect them on a Utah-issued card they can avoid showing. Nobody weighed that statewide cost against that narrow benefit: the bill states no objective and requires no reporting, the fiscal note estimated no business costs although state law required it to, and the amendment that set the checkpoint’s scale arrived on the session’s final day, after the last fiscal note and without a committee hearing.
Retrospective note: this brief is written as it would have appeared at signing on March 27, 2025, from the bill text and the legislative record available that day. A postscript documents what happened after the law took effect.
Legal Impact Assessment
This is not a litigation forecast. It is a vulnerability audit — the governance equivalent of inspecting source code for security flaws. We are not predicting whether vulnerabilities will be exploited. We are identifying that they exist and what they expose.
The federal executive-power toolkit (Youngstown, the major questions doctrine) does not apply to a state statute. The bill’s delegations are bounded. The Alcoholic Beverage Services Commission writes rules for an electronic verification program whose display fields, permitted uses and retention period the statute fixes, and the Driver License Division sets the identifier’s format. These are not the bill’s legal weak points. Its weak points are its title, a criminally enforceable verification duty that examination often cannot perform, the designation and its public marker, and a change in the scale of data collection.
L1. The title does not express the bill’s largest operative change
Utah’s constitution requires that a bill contain one subject, “clearly expressed in its title” (Utah Const. art. VI, § 22). H.B. 437’s short title is Interdicted Person Amendments. Its long title describes designation, court discretion, training, the marked card and voluntary marking. It never mentions verifying proof of age. Yet the bill’s broadest effect falls on people who are not interdicted: removing the 35-or-younger qualifier from §32B-1-407, the statute titled “Verification of proof of age by applicable licensees.” The Highlighted Provisions were identical in the 3rd Substitute, before the floor amendment, and in the enrolled bill, after it. They describe a bill that never contained the floor amendment.
The long title counts. Gregory v. Shurtleff, 2013 UT 18, ¶58, holds that a bill’s long title “is its title for purposes of Article VI, Section 22.” The test is whether the title would “fairly apprise the legislators and the public of the subject matter of the legislation” without misleading them (¶55). That cuts both ways here. The long title never mentions eliminating the 35-or-younger threshold or making electronic verification universal. But it does say the department will train sellers to verify interdiction status “including inspection of the person’s identification card or driver license,” and universal checking can fairly be called an enforcement mechanism for interdiction rather than a separate subject. Utah applies the requirement liberally and does not require a title to list every operational consequence.
The strongest version of the argument is inadequate disclosure of a substantial expansion, not a missing subject. The risk is low. Its value is as a record, not a cause of action: anyone reading the bill’s own description would not learn that it made universal ID scanning the law at every bar in the state.
Textual Finding: High | Litigation Risk: Low
Characterization: Structurally Problematic but Legally Sound.
L2. A duty to verify that examining most IDs cannot fulfill
New §32B-4-405(2) provides that “prior to any sale or furnishing of an alcohol product, a person shall verify whether the person is an interdicted person through examination of” a Utah license or ID card, “or proof of age issued by another state or country.” New §32B-5-306(1)(b) imposes the same duty on every retail licensee, covering minors as well. The interdiction identifier exists only on Utah-issued cards (§53-3-102(25), §53-3-236, §53-3-805(6)). Other states’ licenses, military IDs and passports, all valid “proof of age” under §32B-1-102, cannot carry it.
The duties carry real penalties. A violation of Chapter 4, which includes §32B-4-405(2), is a class B misdemeanor unless the title says otherwise (§32B-4-304(1)). That means up to six months in jail for an individual clerk. Licensees also face administrative discipline. §32B-5-306(1)(b) is not itself a crime, but the same conduct can violate §32B-4-405(2). The “known interdicted person” limit in subsection (1) does not qualify the separate examination duty in subsection (2): failing to examine is its own violation, even if the buyer wasn’t interdicted. Where a statute names no mental state, Utah’s default requires intent, knowledge or recklessness unless strict liability is clearly intended (§76-2-102).
Vagueness turns on which reading a court adopts. Utah’s test asks whether ordinary people get fair notice and whether the law invites arbitrary enforcement (State v. MacGuire, 2004 UT 4, ¶¶13–14). Read as a duty to examine the credential for whatever it shows, the provision is administrable. Read as a duty to determine interdiction status, it demands something examining a passport or another state’s license cannot deliver. Prosecuting a clerk who genuinely examined an acceptable unmarked credential would present the strongest fair-notice challenge. A court would likely avoid it by adopting the first reading. Ineffectiveness alone is not unconstitutional vagueness.
Textual Finding: High | Litigation Risk: Medium (as applied to a seller who examined an unmarked credential; a facial challenge is weaker)
L3. The designation: punitive placement, retroactivity, and a public marker on a general-purpose ID
Ex post facto. The designation sits in the sentencing statutes as an element of the sentence. For a first extreme DUI, the text reads: “As part of any sentence…the court shall…designate the individual as an interdicted person” (§41-6a-505(1)(a)(x)). Similar mandatory language appears in §41-6a-505(5)(a)(x) and (9)(e), and in §76-5-102.1(6) and §76-5-207(8) (”A judge imposing a sentence under this section shall designate…”). It is discretionary for DUI convictions that do not amount to extreme DUI (§41-6a-505(3)(b)(v), (7)(b)(iv)).
The bill gives an effective date but does not say whether the triggering event is the offense, the conviction or the sentencing. Utah presumes Code provisions are not retroactive unless expressly declared so (§68-3-3), which supports prospective application. It doesn’t settle the question for an offense committed before January 1, 2026 and sentenced after.
Applying the designation to such an offense would invite a challenge under Utah Const. art. I, § 18 and U.S. Const. art. I, § 10. Under Smith v. Doe, 538 U.S. 84 (2003), the question is first whether the legislature intended punishment, then whether the scheme is punitive in purpose or effect. Several features point toward punishment:
the designation is imposed as part of the sentence;
the judge sets its length;
it tracks probation;
it brings license surrender and a mandatory public marking.
State v. Trotter, 2014 UT 17, ¶¶29–30, treated sex-offender registration as a collateral consequence partly because it was “beyond the control of the trial court.” This designation is the opposite: the trial court controls it. The state has a real answer: restricting alcohol access serves safety and rehabilitation, and placement in a sentencing statute is evidence of intent, not proof of it.
One caution: interdiction itself, and the ban on purchase by interdicted persons (§32B-4-413, 2010), predate this bill. The ex post facto comparison has to isolate what is new, which is the mandatory designation and the marked card. Prior discretion to impose a restriction does not license making it mandatory retroactively (Lindsey v. Washington, 301 U.S. 397 (1937)).
Textual Finding: Medium | Litigation Risk: High if applied to offenses committed before January 1, 2026; whether the scheme is punitive is contested
The identifier on a general-purpose ID. A designated person must surrender their license or card (§41-6a-509(12); §76-5-102.1(6)(b); §76-5-207(8)(b)). Their replacement shows “No Alcohol Sale” and “a prominent red stripe on the front” (§53-3-102(25)) to everyone who sees the card for any reason: an employer, a landlord, a bank, a pharmacist, a poll worker checking voter ID.
The strongest theory is compelled speech. Two courts have struck down conspicuous sex-offender markings on state IDs: State v. Hill, 341 So. 3d 539, 552–53 (La. 2020), and Doe 1 v. Marshall, 367 F. Supp. 3d 1310, 1323–27 (M.D. Ala. 2019), as applied. Both are persuasive authority, not binding in Utah. The Tenth Circuit’s framework for compelled messages on state-issued documents is Cressman v. Thompson, 798 F.3d 938, 949–51 (10th Cir. 2015), where the plaintiff lost. Utah has real distinctions:
“No Alcohol Sale” states a functional restriction aimed at sellers, and is less stigmatizing than “sex offender.”
The same marking appears on voluntarily obtained cards.
Utah does not require designees to carry the card at all times.
Due process is the second theory. State v. Briggs, 2008 UT 83, separated publishing the fact of a conviction from publishing an implication of current dangerousness, and required process for the latter. Wisconsin v. Constantineau, 400 U.S. 433 (1971), struck down posting a person’s name as barred from buying alcohol without notice or a hearing. Here the designation comes out of a sentencing proceeding, which supplies the process Constantineau found missing. A claim is strongest where the designation is erroneous, misapplied, or still displayed after it expires. A hearing on a fact the statute makes irrelevant is not required (Connecticut Dept. of Public Safety v. Doe, 538 U.S. 1 (2003)).
Textual Finding: High | Litigation Risk: Medium (compelled speech is the strongest theory; due process depends on the facts)
L4. Universal electronic capture of bar patrons’ identity data
Before H.B. 437, §32B-1-407 required electronic verification only for patrons appearing 35 or younger. After it, the statute covers every patron entering a bar or tavern, and every purchase in a dining club or a restaurant dispensing area. The statute sets the maximum a scanner may display: name, age, ID number, birth date, gender and document status. It limits use of the data to verifying proof of age, sets a seven-day retention period, bars disclosure, and makes the data a licensee record (§32B-1-407(5)–(6)). Those constraints are sound (see What This Brief Gets Right). The change is one of scale: a mechanism designed for a subset becomes a register of everyone.
Privacy statutes cover less than one might expect. The Utah Consumer Privacy Act reaches only businesses with at least $25 million in annual revenue that also process large volumes of consumer data (§13-61-102(1)). Most bars fall outside it, and it gives no private right of action (§13-61-305). The Protection of Personal Information Act applies more broadly to name-plus-license-number data, requiring reasonable protection and secure disposal. It too creates no private action (§13-44-301(2)).
The constitutional exposure is government access, not collection. Private ownership of the scanner does not end the inquiry. When the state compels a private search, the Fourth Amendment can apply (Skinner v. Railway Labor Executives’ Ass’n, 489 U.S. 602 (1989)). Utah’s own art. I, § 14 protects some records held by third parties, such as bank records (State v. Thompson, 810 P.2d 415 (Utah 1991)). Licensee records are open to department inspection and audit (§32B-5-302(3)). City of Los Angeles v. Patel, 576 U.S. 409 (2015), struck down a hotel-registry inspection law because owners had no opportunity for review before complying. Alcohol’s status as a closely regulated industry strengthens the state’s position, but it does not make every demand for identifiable patron logs reasonable.
Retention is not as short as the seven-day rule suggests. §32B-5-302(2) requires relevant records to be kept two years after notice of certain dram-shop or DUI proceedings. Nothing in the bill says how that interacts with the seven-day rule. Identifiable patron records may not always disappear after a week.
Textual Finding: High | Litigation Risk: Low (ordinary collection) / Medium (a broad government demand for identifiable patron logs)
L5. Foreign visitors: the definition decides who can be served
“Proof of age” (§32B-1-102(97)) includes passports but not foreign driver licenses. Its driver license clause covers only licenses issued under a U.S. state’s law or by the U.S. Department of State. With verification now universal at bars, and examination required before every sale everywhere, a foreign adult carrying only a home-country license cannot be lawfully served. Nothing in the bill chose that result. It follows from importing a definition built for age verification into a universal check.
A challenge would be weak. The rule distinguishes documents, not nationalities: a foreign visitor can use a passport, and a U.S. citizen holding only a foreign license meets the same wall. Document reliability and verifiability supply a rational basis. A burden-on-tourism theory under the Commerce Clause would need evidence of a substantial burden on interstate commerce relative to the benefits.
Textual Finding: High | Litigation Risk: Low
Structural Analysis
What does this do to the architecture of governance, incentives, and accountability — regardless of the legal outcome?
The design equation
Start with what the bill is trying to do, because the rest of this analysis follows from it. H.B. 437 enforces a ban on specific people at the point of sale. That choice has a necessary consequence: a seller cannot pick out a designated person without identifying every buyer. Universal carding is not a drafting accident. It is what the goal requires. Rep. Eliason’s introduced bill required it from the first version: an ID examination before every sale, by every seller (§32B-4-405(2)) and every retail licensee (§32B-5-306(1)(b)), with no age exception.
Read that way, the Senate floor amendment that struck “who appears to be 35 years of age or younger” from §32B-1-407 follows the same logic. Designated people can be any age, so a check that skips everyone who looks over 35 has a hole in it. It is not an age rule riding an interdiction bill (see Flag 3).
The question the bill never answers is whether what the checkpoint can deliver is worth what it costs.
The cost side scales with every alcohol transaction in the state. It takes the form of seller time, door staff, equipment, turned-away customers and server exposure. It falls almost entirely on businesses and customers, off the state budget. State law required the fiscal note to estimate “anticipated direct expenditure by Utah residents and businesses, including the unit cost, number of units, and total cost” (§36-12-13(2)(c)(iv)). The 3rd Substitute’s note estimated only the fees designated persons would pay, nothing for businesses, and found the bill “likely will not change the regulatory burden for Utah residents or businesses.” Nobody measured the costs.
The benefit side is capped three ways.
The target is small. The fiscal note’s fee projection implies roughly 20 to 100 designees a year (A3).
Detection works on one kind of card. The “No Alcohol Sale” mark exists only on Utah-issued licenses and ID cards. The introduced bill required examining only Utah cards. The 1st Substitute extended the duty to “proof of age issued by another state or country,” so the Legislature knowingly required clerks to examine documents that cannot carry the mark (L2, A4).
The target chooses which card to show. A designated person must surrender a Utah card. Nothing requires them to obtain a marked one, and nothing in the checkpoint detects them if they present a passport, a military ID or another state’s license (Flag 5). That doesn’t make the purchase legal. An interdicted person has been barred from buying or possessing alcohol since 2010 (§32B-4-413), a class B misdemeanor (§32B-4-304(1)). The bill didn’t create the prohibition. What it adds is a way for sellers to detect violations, and the detection works only on the one card a violator can choose not to show.
The benefit therefore depends on the cooperation of the people the law targets. The cost does not.
That also means the checkpoint fails first for exactly the people it most needs to stop. A designee who intends to drink can show a passport, and the server has no way to know. Even a perfect count of refused sales would mostly measure the cooperation of designees least likely to cause harm. And stopping purchases is not the goal anyway; it is a means. The goal behind interdiction is fewer people hurt by designated offenders drinking to excess. The bill writes down neither the goal nor the means, only the checkpoint. Only a measure of that harm could show whether the checkpoint, or interdiction itself, accomplishes anything (Recommendation 3).
What the scanning adds. The every-sale examination duty already covered every retail licensee, bars included. So the floor amendment’s marginal effect was twofold: every patron is carded at the bar door rather than only at the point of sale, and every ID at bars, dining clubs and restaurant dispensing areas is electronically scanned. The scan cannot see the red stripe. On our reading of §32B-1-407(5), it is also not permitted to act on it: its display list contains no restriction field, and its data may be used only “for purposes of verifying proof of age.” (The display list includes the proof of age’s “status,” which could arguably carry a restriction; nothing in the bill says so.)
What the scan plausibly adds is a compliance record. The data is retained for seven days and is “considered a record” for purposes of the retail licensee operational requirements in Title 32B (§32B-1-407(6)(b)). That record proves an ID was presented. It does not prove anyone looked for the stripe. The text never says this is the scan’s purpose. A future connection is half-built: the Driver License Division must make card restrictions scannable “as soon as practicable” (§53-3-104(17)). The fiscal note also budgets for adjusting the scanners at the 51 state liquor stores to “identify the new licenses.” But the bill bars bar and restaurant scanners from using that capability (A2). Nothing in the text connects universal scanning to interdiction.
The alternatives were in the same statute. The DUI sentencing section H.B. 437 amends already directs courts toward tools that act on the offender rather than on every seller. It orders an ignition interlock unless the court finds one unnecessary (§41-6a-505(1)(a)(ix)) and permits a 24-7 sobriety program (§41-6a-505(1)(b)(ii)). Point-of-sale interdiction has real advantages over those tools: it reaches buying, not just driving, and needs no monitoring infrastructure. A legislature could reasonably choose it. The defect is not the choice. It is that no record shows the choice was compared, priced or scaled to what the checkpoint could detect.
Why the trade was never weighed. The bill states no objective (Flag 2, A1), so there was no stated benefit to weigh. The fiscal note, though required to estimate costs to businesses, estimated none. And the amendment that set the checkpoint’s scale bypassed the committee hearing and the fiscal note (Flag 8). Each of those findings stands on its own. Together they explain how a statewide cost was attached to a narrow, avoidable benefit without anyone being required to notice.
Textual Finding: High (the universal duty, the coverage limits, the scan’s statutory limits, the sentencing alternatives) / Medium (the scan’s compliance-record function; the population estimate) | Structural Significance: High
Structural flags
1. Power Concentration — ABSENT. The bill does not concentrate decision-making authority or replace a governance pathway. If anything it distributes enforcement outward, onto every person who sells alcohol. Textual Finding: High | Structural Significance: Low
2. Accountability Gaps — CRITICAL. This is the flag that drives the brief. H.B. 437 contains no findings, purpose or intent section, and no stated outcome anywhere in its text. Its only purpose-adjacent language is the long title, which describes what the bill does and never what it is for. It requires no reporting: no count of designations, marked cards issued, refused sales, or compliance checks for interdiction. The fiscal note’s performance-evaluation section, prepared under the Legislature’s own joint rules (JR1-4-601), recorded that the bill “does not create a new program or significantly expand an existing program.” That is the finding on which a performance evaluation turns, and it ruled evaluation out.
So when this law runs, no actor will hold information showing whether designated people are being stopped, let alone whether anyone is safer for it. Every outcome will look identical to the institutions responsible: a working interdiction regime and a failing one. The only signals that will reach the Legislature are the ones the design generates without trying: complaints from the people it inconveniences.
Textual Finding: High | Structural Significance: High
3. Bundling — ABSENT. On first reading, the universal scan looks like an age-verification expansion riding an interdiction bill. The amended statute is titled “Verification of proof of age,” and its data is confined to age. But removing the 35-or-younger qualifier follows from the interdiction goal itself: designated people can be any age, so carding only people who look young leaves a hole. The remedy belongs to the predicate. The real defect is different. The bill implements an interdiction rule with age machinery that cannot detect interdiction (The design equation, A2), and its title never says the expansion is there (L1). That is a mismatch between mechanism and goal, not an unrelated remedy.
Textual Finding: High | Structural Significance: Low (as bundling; the underlying mismatch is rated under A2)
4. Vague Enforcement — PRESENT. “Verify whether the person is an interdicted person” (§32B-4-405(2); §32B-5-306(1)(b)) is undefined and has no standard for documents that cannot carry the identifier (L2). The machine-readable restriction the Driver License Division must build carries no deadline: it is due “as soon as practicable” (§53-3-104(17)). The division’s authority to issue the marked card is permissive: it “may accept an application” and “may provide” the card (§53-3-236(1); §53-3-805(6)(a)). The bill says what the division may not issue to a designated person: an unmarked card (§53-3-236(2); §53-3-805(6)(c)). It never says what follows if the person holds no Utah card at all.
Textual Finding: High | Structural Significance: High
5. Perverse Incentives — PRESENT. Zombie Emergency Trap — ABSENT. The design rewards a designated person for never obtaining the card the scheme depends on. Someone who surrenders their license and never applies for a marked one carries no identifier. They can present a passport, a military ID, or a license from another state, and pass every examination the bill requires. The purchase is still a crime (§32B-4-413). What the design rewards is concealment, not legality. Someone who complies and obtains the marked card pays a $7 identifier fee (§53-3-105(40)), plus the duplicate-license fee where applicable (§53-3-236(6)), and is the only designated person the check can catch. The scheme costs the compliant and is invisible to the evasive. No zombie loop exists: resolving the problem would not remove the bill’s legal basis.
Textual Finding: High | Structural Significance: Medium
6. Third-Party Incentive Gaps — PRESENT. Detection depends entirely on intermediaries: servers, clerks and door staff. Their liability is still keyed to selling to a known interdicted person. Nothing in the design aligns any party’s obligations with actually catching designated people. The division’s issuance is permissive and its scan capability is undated. The commission’s scanner specification is confined to age. Licensees have their liability capped by knowledge, while their workload rises sharply.
Textual Finding: High | Structural Significance: High
7. Sunset Provisions — ABSENT. No sunset, review date, or reauthorization trigger. Combined with Flag 2, this makes the regime permanent by default and invisible to evaluation by design.
Textual Finding: High | Structural Significance: Medium
8. Preemption of Oversight — PRESENT (inadvertent, procedural). The bill reduces no oversight mechanism in law. But the provision that sets its footprint, universal scanning, bypassed every review stage the Legislature has. The record:
The House committee’s hearing on February 12 covered the 1st Substitute, which did not touch §32B-1-407.
The §32B-1-407 amendments first appeared in the 2nd Substitute, adopted on the House floor on February 20. That version kept the 35-or-younger qualifier and added a hardcoded scanner requirement.
The Senate committee recorded the bill as “Not Considered” on February 28 and returned it to Rules on March 4.
The 3rd Substitute’s fiscal note (March 6, 12:07 p.m.) concluded the bill “likely will not change the regulatory burden for Utah residents or businesses.” Its section on individuals and businesses estimated only the fees designated persons would pay, although §36-12-13(2)(c)(iv) required an estimate of direct expenditure by businesses. It said all this even though that version already required the verification technology used by bars, taverns, dining clubs and restaurants to recognize “every state’s unique hidden security features.”
Senate Floor Amendment 1, dated March 6 at 5:08 p.m. and proposed by Sen. Stevenson, struck “who appears to be 35 years of age or younger.” The Senate adopted it on March 7, the 45th and final day of the session, during second and third readings under suspension of the rules. The House concurred that afternoon.
No committee heard any version containing the §32B-1-407 changes. No fiscal note was prepared after the floor amendment that made scanning universal. Nothing in the record suggests this was designed to avoid review. The flag is marked inadvertent. The structural point stands either way: the Legislature’s end-of-session process let the largest change in the bill’s real-world footprint pass with no committee review and no new fiscal note.
Textual Finding: High (official action history and amendment text) | Structural Significance: High
9. Inter-Agency Cannibalization — ABSENT. Each function goes to the domain agency: designation to courts, identifiers to the Driver License Division, verification rules and training to the Alcoholic Beverage Services Commission and its department. The defect is not misrouting but an unconnected interface between the agencies (see Abstraction Layer Analysis A2).
Textual Finding: High | Structural Significance: Low
10. Exemptions Architecture — ABSENT (minor). The discretionary channels are narrow and bounded. They are: a commission-established process for proof of age that cannot be electronically verified (§32B-1-407(4)(b)(ii)); the existing medical exception (§32B-4-405(3)); and a court’s “good cause” adjustment of the designation period. None creates an opaque relief system.
Textual Finding: High | Structural Significance: Low
Incentive mapping
The policy’s beneficiaries are clear: people on Utah roads, if designated drivers are stopped from buying. The structure’s beneficiaries are different. Legislators get visible action on extreme DUI at a fiscal cost to the state of about half a million dollars, mostly one-time Driver License Division programming. The costs that matter fall off-budget, on businesses and customers, as door time, equipment, turned-away patrons and staff exposure. State law required the fiscal note to estimate direct expenditure by businesses (§36-12-13(2)(c)(iv)), but the note estimated none for them. Its own disclaimer adds that fiscal notes “do not measure a bill’s benefits or non-fiscal impacts like opportunity costs, wait times, or inconvenience.” The design also creates demand for ID-scanning equipment able to meet a statutory hidden-security-feature standard. Nothing in the record shows anyone sought that outcome, and the analysis does not depend on it. The mechanism creates the demand whoever proposed it.
The cost distribution is classic. The burden is diffuse, spread across every alcohol transaction in the state. Its source is invisible to the people bearing it, who will read it as an age rule. And it is hard to trace to any decision: no committee discussed it, and the amendment that created it was adopted without a roll-call vote. The bill as a whole then passed on recorded votes. Those are the conditions under which a burden is politically sustainable. They also mean that when it stops being sustainable, the reaction will be aimed at the burden, not the design.
Textual Finding: Medium | Structural Significance: Medium
Failure modes
Measurement void. The design produces no information about its subject, and its subject is itself only a means. The real objective, less harm from designated offenders drinking, sits two steps beyond anything the bill observes. When a system cannot observe its own objective, it can only be corrected through channels it does observe. This is a design output, not an implementation lapse: no reporting duty means no data, whoever administers the law.
Textual Finding: High | Structural Significance: High
Proxy substitution. The universally visible mechanism is age verification. Because the bill states no other objective, the public, licensees and enforcers will treat it as an age law, since age is the only thing the visible mechanism checks. Complaints will be about age and inconvenience. They will be correct readings of the only available signal.
Textual Finding: High | Structural Significance: High
Coverage collapse. The identifier works on one class of document; the check runs on all classes. Effectiveness depends on designated people choosing to present the one document that catches them.
Textual Finding: High | Structural Significance: High
Second- and third-order effects (predictable at signing)
Second order. Door and counter time rises at every licensee. Patrons who are not carrying ID are turned away. Foreign visitors without passports cannot be served (L5). Licensees buy or upgrade scanners to meet the hidden-security-feature mandate. The department builds training on spotting a marker that will be scarce for years: the first marked cards can only follow sentences imposed on or after January 1, 2026, the day the universal checks begin (A3).
Third order. The correction will arrive through the only channel the design feeds: complaints about burden. A legislature revising this law will have abundant evidence of inconvenience and no evidence about interdiction. It will optimize the variable it can see. The predictable result is a rollback along the burden axis that also removes interdiction checks, with no one able to say whether that trades away anything real. Beyond this law, the episode teaches the public that safety measures built this way are inconvenience generators, and it leaves universal ID-scanning infrastructure installed and normalized for whatever purpose next uses it.
Textual Finding: Medium | Structural Significance: High
Abstraction Layer Analysis
Read from the enrolled text and the four prior versions.
A1. No policy layer at all. Well-architected legislation states the goal at the statutory layer and delegates the mechanism. H.B. 437 is all mechanism: designation, surrender, card format, examination, scanning, fees. The API contract has implementations and no interface definition. Every downstream question — does scanning serve the goal, should the 35 rule survive, is a passport check adequate — has no answer in the text, because the text never says what the mechanism is for.
Textual Finding: High | Structural Significance: High
A2. Two layers built to different specifications, with no interface between them. The Driver License Division must, “as soon as practicable,” make license and ID restrictions readable “through an electronic scan” (§53-3-104(17)). The Commission’s verification program specifies what licensee scanners may display: name, age, ID number, birth date, gender, and “the status and expiration date of the individual’s proof of age.” Restrictions are not on that list. The program limits use of the data to “verifying proof of age” (§32B-1-407(5)(a)). Even after the division ships a machine-readable restriction, the licensee layer is barred, by display list and by use limitation, from acting on it. The bill writes both specifications and connects neither.
Textual Finding: High | Structural Significance: High
A3. The layers switch on in the wrong order. The universal examination and scanning duties, the designation authority, and the marked card all take effect on the same date, January 1, 2026. Designations attach at sentencing. So on the day the checks begin, the marked-card population is close to zero: voluntary applicants plus anyone sentenced that day. It grows only as new sentences accrue. The bill’s own fiscal note implies how slowly. It projected 7 identifier fee and a $30 one-time duplicate-license fee. That implies roughly 20 to 100 designees a year. The note itself states no population estimate. The checkpoint runs at full scale from day one on a target population measured in dozens.
Textual Finding: High (simultaneous effective date) / Medium (population inference from fiscal-note arithmetic) | Structural Significance: High
A4. Coverage audit: an age definition imported into an interdiction check. The verification duties operate on “proof of age” (§32B-1-102(97)), a definition built to establish a birth date. For age, it correctly includes passports, military IDs and other states’ cards. For interdiction, none of those can carry the identifier. And by excluding foreign driver licenses, the same definition becomes, under universal verification, a rule that foreign visitors need a passport to be served (L5). The import carries the right population for one purpose and the wrong one for the other.
Textual Finding: High | Structural Significance: High
A5. Undefined term at the enforcement trigger. “Verify whether the person is an interdicted person” is the operative duty and is undefined (L2, Flag 4). Two drafting artifacts sit beside it. Both new subsections say “alcohol product” where Title 32B’s defined term is “alcoholic product.” And “a person shall verify whether the person is an interdicted person” uses the same word for the seller and the buyer. Neither will matter in practice. Both are signs that the duty was not drafted against the title it sits in.
Textual Finding: High | Structural Significance: Medium (the undefined duty) / Low (the artifacts)
A6. Technology specification hardcoded in statute. New §32B-1-407(5)(b) requires that the verification program include “technology that recognizes every state’s unique hidden security features located on state issued identification cards.” That is an implementation detail: vendor capability changes, states redesign cards, and the requirement may not be satisfiable as written. Correcting it now requires new legislation. It belonged in the commission’s rulemaking, which the same section already directs.
Textual Finding: High | Structural Significance: Medium
A7. Documentation drift: the bill’s description does not match the bill. The long title and Highlighted Provisions were not updated for either §32B-1-407 change: the 2nd Substitute’s addition or the floor amendment’s deletion. A reader of the bill’s own summary, including any legislator voting on concurrence, would see an interdiction bill. The summary is not enacted law. It is the interface most people read.
Textual Finding: High | Structural Significance: Medium
A8. Amendment drift: interdiction coverage is coupled to the age definition. Because the verification duties run on “proof of age,” any future change to that definition, made for reasons that have nothing to do with interdiction, silently changes interdiction coverage. Adding a new acceptable document widens the gap if the document cannot carry the identifier. A foreign license, or a digital or mobile ID with no red stripe to see, would be examples. The bill couples its enforcement to a definition it does not control and gives no instruction for keeping the two aligned.
Textual Finding: High | Structural Significance: Medium
A9. The standard defect-catching stages were bypassed. For a state legislature, those stages are the committee hearing and the fiscal and regulatory-impact note. The defining provision passed through neither (Flag 8). The one note that was prepared, on a version that already mandated statewide scanner capabilities, found no change in regulatory burden.
Textual Finding: High | Structural Significance: High
What This Brief Gets Right
These provisions are sound and should survive any revision.
Voluntary self-interdiction. Any person may choose to carry the marked card, with a 30-day lock before reversing (§53-3-236(3); §53-3-805(6)(b)). It is a self-exclusion tool of the kind gambling regulators use. It respects autonomy, costs the state almost nothing, and gives people in recovery a mechanism they control.
Textual Finding: High | Structural Significance: Medium
A court designates, not an agency. Designation happens at sentencing, by a judge, bounded by the probationary period, with a good-cause adjustment. That anchors a significant restriction in an adversarial proceeding with counsel and appeal, rather than in administrative discretion.
Textual Finding: High | Structural Significance: Medium
Data minimization in the verification program. The display fields, the use limitation, the seven-day retention limit and the nondisclosure rule (§32B-1-407(5)–(6)) predate this bill. The bill preserves them while expanding coverage. They are the right constraints, and any expansion of scanning should keep them.
Textual Finding: High | Structural Significance: Medium
A low-tech identifier at the right layer. “No Alcohol Sale” and a red stripe can be read by any clerk with no equipment, network or vendor. For a binary flag on a physical card, a human-readable mark is the most robust design available, and it does not depend on the scanner layer that the bill fails to connect (A2).
Textual Finding: High | Structural Significance: Medium
The right direction on machine-readability. §53-3-104(17) recognizes that a restriction printed on a card should also be readable electronically. It is unsynchronized and undated (A2, Flag 4), but it points at the correct end state.
Textual Finding: High | Structural Significance: Low
The Architecture Question
What does this do to the long-term architecture of democratic accountability? What does it enable next?
While it runs. H.B. 437 converts ID verification from a targeted, judgment-based check into a universal one, and builds a data-capture step into entry at every bar in the state. It does so with no channel back to the Legislature about whether its stated subject, interdiction, is achieved. The system gains reach and loses sight at the same time.
The template. Two templates are established, and both are reusable. The first is restriction-by-card-marker plus universal checkpoint: any future status a legislature wants enforced at the point of sale can ride the same infrastructure, with the same coverage gap and the same absence of measurement. The second is procedural: the decisive scope change made by floor amendment on the final day, after the fiscal note, outside any committee. That path is available to every future sponsor of either party, on any subject. A template that works without anyone choosing to abuse it is more durable than one that requires bad intent.
What enactment fails to fix. The bill responds to a real problem: extreme DUI offenders who keep buying alcohol. But it does not create the conditions under which anyone could learn whether the response works. When the law is revised, as laws of this shape are, the revision will be informed by inconvenience, not by results.
The trajectory. Applied repeatedly, this pattern produces a legislature that legislates by mechanism, corrects by complaint, and never accumulates evidence about objectives. Each cycle adds or removes burden. None adds knowledge. Over five to twenty years the statute book fills with mechanisms whose purposes exist only in the memories of their sponsors. Those memories leave office, and the next legislature inherits machinery it cannot evaluate.
Structural gap statement. The Utah Legislature requires a fiscal note to estimate what a bill will cost the treasury and what it will cost residents and businesses. Nothing enforces the second estimate, and nothing forces either one to be redone when a floor amendment changes the bill. It has nothing requiring a bill’s cost to the public to be weighed against what its mechanism can actually deliver, none requiring a bill to state what it is supposed to accomplish, none requiring the state to find out whether it did, and none that forces a scope-expanding floor amendment back through the cost estimate and committee review that the rest of the bill received.
Recommendations
The question is not only whether this was done. It is what functional design would look like instead.
1. Price the checkpoint against what it can detect, and compare it to the alternatives, before scaling it. (Addresses The design equation.) Before making ID checks universal, the Legislature should have a written comparison of three things:
point-of-sale interdiction;
the offender-directed tools already in §41-6a-505, namely ignition interlock and 24-7 sobriety programs;
a narrower checkpoint: visual inspection of Utah cards wherever ID is already checked, with no universal scanning.
The comparison should include private compliance costs, which §36-12-13(2)(c)(iv) already requires the fiscal note to estimate, and an honest statement of what each option can detect. Only an outcome measure can settle which option actually reduces harm (Recommendation 3). Without one, this comparison stays a matter of judgment, and judgment is how the bill got built. Buying is already a crime for a designated person (§32B-4-413), so the question for the checkpoint is detection, not prohibition. If the checkpoint is kept, put more of the detection burden on the offender and less on every seller. One way is a probation condition barring the designee from using any other credential to obtain alcohol. A violation would then be a probation violation, with supervision and swifter consequences, rather than a misdemeanor that someone has to catch and prosecute. Any such condition should be narrow: alcohol purchases only, tied to the supervision period, and leaving unrelated uses of a passport or other ID untouched.
2. State the objective in the operative text, and define the verification duty against it. (Addresses Accountability Gaps, A1, A5, A8, and the specific ambiguity in §32B-4-405(2).) Amend the bill to state the objective in operative text, in terms of harm rather than mechanism: interdiction exists to reduce alcohol-related harm, especially impaired driving, by people a court has designated. Keeping alcohol from them is a means to that end, and point-of-sale checking is one way of doing it, to be judged against the alternatives in Recommendation 1. Then define “verify” honestly. For a Utah-issued license or ID, verification means visual inspection for the interdiction identifier. For any other proof of age, no interdiction verification is possible, and the statute should say so rather than deem it performed. Write the interdiction duty against its own list of covered documents rather than the general “proof of age” definition, so that future changes to the age rules do not silently change it. That converts an undefined duty into a stated one with a stated coverage limit, which is the precondition for anyone measuring it.
3. Measure the outcome, report the mechanism, and set a review date. (Addresses Accountability Gaps, Sunset, Measurement void.) Direct an annual report to an interim committee with two tiers.
The outcome, as the primary measure. Among designated people, track repeat impaired-driving arrests and convictions, alcohol-involved crashes, and other alcohol-related offenses during and after designation, against a comparison group. The cleanest comparison is a natural experiment the effective date creates: offenders sentenced for extreme DUI just before January 1, 2026, who could not be designated, against those sentenced just after, who had to be. Courts and public-safety agencies already record arrests, convictions and crashes, so most of the data exists. The work is linking it to designation status.
The mechanism, as diagnostics. Designations ordered, marked cards issued, voluntary applications, sales refused for interdiction, and interdiction-related compliance findings. These do not show whether the law works. They show where the chain broke if it doesn’t: few cards issued, few refusals, or designees buying with other ID.
Set a review in the 2028 interim, with the universal verification provisions expiring unless reauthorized on the outcome record, not the output counts. A law can post excellent output numbers while preventing no harm at all, and this design makes that the likely result. Without an outcome measure, no future legislature can tell whether H.B. 437 did anything.
4. Synchronize the layers before making the checkpoint universal. (Addresses A2, A3, Coverage collapse.) Either amend §32B-1-407(5) to allow display of a restriction flag and permit its use for interdiction, or drop the universal scanning mandate until the identifier can be scanned. Tie the effective date of universal verification to the division certifying that machine-readable restrictions exist. Until then, the existing 35-or-younger scanning rule, plus visual inspection of Utah cards for the stripe, delivers everything the universal scan can deliver for interdiction, at a fraction of the burden.
5. Move technology specifications to rule. (Addresses A6.) Strike §32B-1-407(5)(b) and direct the commission to set and update scanner standards by rule, under the existing rulemaking act. Standards will change faster than sessions.
6. Close the end-of-session path. (Addresses Flag 8, A7, A9, and the structural gap.) The Legislature should amend its joint rules so that any floor amendment changing the population a bill regulates triggers two things before the final vote: an updated fiscal and regulatory-impact note, and a conforming change to the long title. This is a rule change, not a statute. It costs one note. It would have put the words “applies to every patron of every bar” in front of the House on March 7.
7. Preserve what works. Keep voluntary self-interdiction, court-anchored designation, the data-minimization constraints and the human-readable identifier. They are the sound core of the bill, and the corrections above are built around them.
These are the kinds of recommendations a Governance Design Agency would make routinely: state the objective, instrument the mechanism, and keep the decisive change inside the review process.
Postscript
What happened after March 27, 2025.
The law took effect January 1, 2026. The Department of Alcoholic Beverage Services’ guidance stated what the text implied. ID scanners “may not catch whether the ID is that of an interdicted individual.” Employees “should rely on the visible ID check.” And: “The ID notation is not retroactive, therefore we can expect to see very few interdicted IDs.” Passports “will not be adjusted to include an ‘interdicted’ notation” and “remain an acceptable form of ID.” (A2, A3 and A4, confirmed by the implementing agency.)
Complaints arrived on the axis the design predicted. Older patrons and regulars pushed back. Out-of-town patrons who had left their ID at the hotel could not order a drink. Restaurants said the rule was “impacting their business.” The complaints were about age and inconvenience, because age was the only thing the visible mechanism checked. (Proxy substitution.)
The sponsor disavowed the operation of his own law. Rep. Eliason told PBS Utah that ID checks for people “clearly over 21” entering a restaurant “was actually never the intent,” and: “That was not to actually verify age.” He described the purpose as keeping alcohol from interdicted extreme-DUI offenders. Universal scanning did come from the Senate floor amendment, not from his drafts. Universal carding did not. His introduced bill already required that every buyer’s ID be examined before every sale, at restaurants and everywhere else (§32B-5-306(1)(b); §32B-4-405(2)). It made no exception for people who are clearly over 21. The carding he disavowed was in his own text from the first version. And the section the bill amended is titled Verification of proof of age and restricts its data to “verifying proof of age.” The intent existed. It was never written into the text, so nothing in the text could hold either the bill or the amendment to it.
The Legislature cut the law back, along the burden axis, within one session, in two overlapping bills.
H.B. 597 (2026), Alcohol Amendments, was sponsored by Rep. Burton. Its Senate sponsor was Sen. Stevenson, who had offered the 2025 floor amendment. It was signed March 13, 2026, and its changes to §32B-1-407 took effect on signature. It made foreign driver licenses valid proof of age effective May 6, 2026 (L5).
H.B. 59 (2026), Identification Verification Amendments, was again sponsored by Rep. Eliason. It was signed March 17, 2026, and its verification changes apply retrospectively to January 1, 2026. Its identifier-card changes took effect May 6. A coordination clause makes its version of §32B-1-407 control over H.B. 597’s.
Between them, the two bills:
restored the 35-or-younger rule at restaurants, now applying anywhere on the premises rather than only in the dispensing area;
kept universal scanning at bars and taverns, and extended it to curbside and drive-through beer pickup;
repealed the every-sale examination duties in §32B-4-405(2) and §32B-5-306(1)(b) (H.B. 59). In their place, state stores, package agencies and off-premise beer retailers must request ID from every purchaser;
dropped the hidden-security-feature mandate (A6). H.B. 597 repealed it, and H.B. 59 replaced it with a barcode-validation standard;
limited the interdiction check, wherever proof of age is verified under §32B-1-407, to “a driver license issued by this state.” H.B. 59 likewise required state stores, package agencies and off-premise beer retailers to check only “a Utah driver license or an identification card.” That is the Legislature writing A4’s coverage gap into law;
limited the surrender requirement to Utah-issued cards, and waived the identifier fee for voluntary applicants (H.B. 59). H.B. 59’s Highlighted Provisions say the court “may” require surrender, but its operative text still says the court “shall.” Once again the bill’s description and the bill disagree (A7).
H.B. 59’s general description: “This bill amends provisions relating to identification verification.” Neither bill contains a purpose statement.
Whether interdiction worked is unknowable. No baseline was set, no measure defined, and no report required, so no data exists against which an effect could be estimated, in either direction. The correction was made without it, as the design guaranteed it would be.
What the outcome failed to fix. Both 2026 bills adjust mechanism, and they move the checkpoint closer to what it can detect. They limit interdiction verification to the Utah card that can carry the mark and drop most of the restaurant burden, while keeping universal checks at bars. But they got there by complaint, not by comparison. Neither bill states what interdiction is for, requires a report, or sets a review. The next revision will again have evidence about burden and none about results.
The vulnerabilities identified above were not predictions. They were observable in the legal architecture at the time of signing. The outcome followed from the design.
Sources
Primary — Utah Legislature, 2025 General Session, H.B. 437
Bill page and action history — status, votes, committee actions, version list
Introduced bill and the comparison of the introduced bill to the 3rd Substitute; the 1st and 2nd Substitutes are linked from the bill page
Senate Floor Amendment 1 (Sen. Stevenson, March 6, 2025)
3rd Substitute fiscal note (March 6, 2025)
Primary — 2026 General Session (postscript)
H.B. 59, Identification Verification Amendments: bill page, enrolled text
H.B. 597, Alcohol Amendments: bill page, enrolled text
Agency and press (postscript)
Utah DABS, New 100% ID Law Begins Jan. 1, 2026 (Dec. 19, 2025; updated Jan. 26, 2026)
Utah DABS, Utah Legislative Changes to the 100% ID Law Requirements – Effective Immediately (Mar. 18, 2026)
Lauren Steinbrecher, Lawmakers Move to Loosen Utah Alcohol Laws, PBS Utah (Feb. 26, 2026)
Ben Winslow, 2026 Utah Legislature on alcohol, FOX 13 (Mar. 7, 2026)
Not reviewed: House floor debate audio (Feb. 20, 2025), Senate floor debate audio (Mar. 7, 2025), and committee minutes. The analysis rests on the enacted text and the official record. The floor audio may contain the stated rationale for Senate Floor Amendment 1.