Bills & Beans

Straight Talk on Coffee, Money, and Liberty

Subscribe

Tag: Myth Check

A widely held claim, tested against a number.

  • Free Speech and Private Platforms: The First Amendment Restrains Government

    Free Speech and Private Platforms: The First Amendment Restrains Government

    In this piece

    1. A private entity can qualify as a state actor in limited circumstances
    2. A company that runs an entire town can count as a state actor
    3. A site that removes protected speech in good faith can claim Section 230’s shield
    4. A platform can hold a First Amendment right in how it curates a feed
    5. An official who coerces a company over speech can face a First Amendment claim
    6. What’s still unsettled
    An old brass ribbon microphone on a short stand against a dark green background, its grille turned slightly off-axis

    A social media platform suspends an account. Someone calls it a First Amendment violation. The amendment restrains government, and in 2019 the Supreme Court said its speech clause reaches only governmental abridgment of speech. Congress wrote a separate statute in 1996, and it shields a site’s good-faith removals, “whether or not such material is constitutionally protected.”

    A private entity can qualify as a state actor in limited circumstances

    The grievance quotes a text. The answer should quote one too. The First Amendment protects “the freedom of speech.” Ratified in 1868, the Fourteenth Amendment applies that clause against the States. The Supreme Court put the result plainly in 2019: “the Free Speech Clause prohibits only governmental abridgment of speech.”

    Justice Kavanaugh wrote that opinion, in Manhattan Community Access Corp. v. Halleck. Its first sentence is the cleanest available. “The Free Speech Clause of the First Amendment constrains governmental actors and protects private actors.” The vote was five to four. The date was June 17, 2019. Halleck concerned public access cable channels in Manhattan. A private nonprofit designated by the City ran them.

    The grievance fixes on a label, censorship. The governing doctrine fixes on an actor. Lawyers call it state action. Halleck gave examples of how a private entity becomes a state actor. It can perform a traditional, exclusive public function. The government can compel it to act, or act jointly with it.

    Advertisement

    A company that runs an entire town can count as a state actor

    The myth’s best precedent is a company-town case. Marsh v. Alabama was decided January 7, 1946. A Jehovah’s Witness set out to distribute religious literature on a sidewalk in Chickasaw, Alabama. The Gulf Shipbuilding Corporation owned the whole town. A deputy of the county sheriff, paid by the company, served as the town’s policeman.

    Justice Black wrote for the Court. Apart from its owner, he wrote, the town had “all the characteristics of any other American town.” A posted company notice barred solicitation without permission. Alabama enforced it with a criminal statute. The Court held that the State’s attempt to punish her for distributing that literature in a company town “cannot stand.”

    Halleck later listed operating a company town among the “very few” traditional, exclusive public functions. Running elections is on the same list. Then the opinion described what ordinary owners do. “Grocery stores put up community bulletin boards. Comedy clubs host open mic nights.” On that reasoning, a comment section sits with the bulletin board, far from a town whose owner paid the policeman.

    A site that removes protected speech in good faith can claim Section 230’s shield

    Congress enacted its own rule for online services on February 8, 1996. It sits at 47 U.S.C. § 230. The heading reads “Protection for private blocking and screening of offensive material.” Subsection (c)(1) is the famous one. “No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.” It names the user alongside the provider, so its protection reaches both.

    Subsection (c)(2) is the moderation clause. It covers voluntary, good-faith action “to restrict access to or availability of material.” The list of what the provider or user “considers to be” objectionable runs from “obscene” to “otherwise objectionable.” Then come eight words. The shield holds “whether or not such material is constitutionally protected.” The statute anticipates sites taking down protected speech, and it bars liability for doing so in good faith.

    The same section draws the shield’s edges. Subsection (e) keeps federal criminal law outside it. Intellectual property and communications privacy law stay outside too. A 2018 amendment carved some sex-trafficking claims and charges out of the rest of the section. The good-faith removal clause still reaches them. State law consistent with the section survives. The text quoted here is the version the House’s online U.S. Code printed as in effect on September 9, 2026.

    A platform can hold a First Amendment right in how it curates a feed

    Florida and Texas legislated in 2021. Both laws curtailed large platforms’ content moderation. The two challenges were decided together. Justice Kagan wrote for the Court in Moody v. NetChoice on July 1, 2024. Its judgment rested on how the lower courts had handled facial challenges, and its reasoning about feeds went well past that.

    Its precedents, the opinion said, hold that “expressive activity includes presenting a curated compilation of speech originally created by others.” Then it took Texas at its word: the objective was “to correct the mix of speech that the major social-media platforms present.” The opinion’s reply was blunt. “But a State may not interfere with private actors’ speech to advance its own vision of ideological balance.”

    Advertisement

    An official who coerces a company over speech can face a First Amendment claim

    A user’s real protection points at officials. On May 30, 2024, a unanimous Court decided National Rifle Association of America v. Vullo. The complaint said a New York financial regulator had pressed companies she oversaw to cut ties with an advocacy group. Justice Sotomayor’s opinion put the rule in a single sentence.

    “Government officials cannot attempt to coerce private parties in order to punish or suppress views that the government disfavors.” The Court held that a complaint plausibly alleging that kind of coercion states a First Amendment claim. The rule is older than the internet. It traces to Bantam Books, Inc. v. Sullivan, decided in 1963. A company curating its own feed holds a right, and an official who coerces it answers to the First Amendment.

    The short version

    What’s still unsettled

    As of September 2026, the Supreme Court has given no merits answer to two platform questions. The first is when an official’s pressure turns a platform’s removal into state action. Murthy v. Missouri, decided June 26, 2024, held that the plaintiffs had not established standing to seek an injunction, and the Court said it lacked jurisdiction to reach the merits. The second is how far Moody’s account of curated feeds reaches other features, since that opinion suggested feeds and direct messages “involve different levels of editorial choice.”

    This is educational content explaining constitutional and statutory doctrine, not commentary on any pending legislation or litigation.

    Article sources

    1. National Archives — The Bill of Rights: A Transcription (1789 Joint Resolution and ratified text, Amendment I)
    2. Cornell Law School LII — Manhattan Community Access Corp. v. Halleck, No. 17-1702 (June 17, 2019), syllabus, opinion of the Court (Kavanaugh, J.) and dissent
    3. uscode.house.gov (U.S. House of Representatives) — 47 U.S.C. § 230, text in effect on September 9, 2026, with source credit and amendment notes
    4. Cornell Law School LII — National Rifle Association of America v. Vullo, No. 22-842 (May 30, 2024), syllabus and opinion of the Court (Sotomayor, J.)
  • Marginal vs. Effective Tax Rate: The Bracket You’re In Isn’t the Rate You Pay

    Marginal vs. Effective Tax Rate: The Bracket You’re In Isn’t the Rate You Pay

    In this piece

    1. The bracket is a boundary, not a label on your income
    2. Walk $160,000 through the bands one at a time
    3. Two effective rates, and most people never say which one they mean
    4. Only the marginal rate prices the next dollar you defer
    5. Payroll tax sits on top and ignores brackets entirely
    6. Twenty-two percent touches only the last $27,000 of this household’s taxable income
    A tall glass cylinder holding four horizontal bands of sand at different depths, running pale gold at the top to near-black at the bottom, lit by a single warm light

    “I’m in the 22% bracket” is the most repeated sentence in American personal finance. It names a boundary, not a bill. Run an illustrative married couple’s $160,000 of wages through the 2026 joint tables and the effective federal income tax rate lands at 13.72% of taxable income.

    The bracket is a boundary, not a label on your income

    A tax bracket is a range of taxable income. The rate attached to it applies only to the dollars that land inside it. Crossing into 22% does not reprice the dollars below the line.

    The 2026 tables set the ranges out plainly. For a married couple filing jointly, 10% covers taxable income up to $24,800. From there, 12% runs to $100,800 and 22% runs to $211,400, per the IRS 2026 inflation adjustment tables.

    The joint tables are not the single tables doubled, either. Joint thresholds do match twice the single figures through the floor of the 35% bracket. There, $512,450 is exactly two times $256,225. Then they diverge: 37% starts at $768,700 for a couple and at $640,600 for a single filer.

    Advertisement

    Walk $160,000 through the bands one at a time

    Start with gross wages of $160,000. Subtract the 2026 standard deduction for a joint return, $32,200. Taxable income is $127,800. That figure, not the $160,000, is what meets the brackets.

    Now take it band by band.

    Band of taxable incomeRateAmount taxed in this bandTax
    $0 – $24,80010%$24,800$2,480
    $24,800 – $100,80012%$76,000$9,120
    $100,800 – $127,80022%$27,000$5,940
    Total$127,800$17,540

    Bracket thresholds and the $32,200 joint standard deduction from IRS Rev. Proc. 2025-32 (published October 9, 2025), tax year 2026. The $160,000 household is illustrative.

    The marginal rate here is 22%. The next dollar of taxable income still falls in the band running to $211,400. Total federal income tax is $17,540. Only $27,000 of the $127,800 was ever exposed to 22%, which is just over a fifth of it.

    Two effective rates, and most people never say which one they mean

    Divide $17,540 by taxable income of $127,800 and the effective rate is 13.72%. Divide the same $17,540 by gross wages of $160,000 and it is 10.96%.

    Both are defensible. Neither is 22%.

    The gap between the bracket and the taxable-income measure is 8.28 percentage points. Measured against gross wages, it widens to 11.04 points. That distance is the entire myth.

    So when someone quotes an effective rate, ask what sits in the denominator. Gross wages give the lower, friendlier number, because the standard deduction is still inside it. Taxable income gives the higher one. Unlabelled, the figure means very little.

    Only the marginal rate prices the next dollar you defer

    There is one job the marginal rate does that the effective rate cannot. It prices the next dollar, and a deduction, a deferral or a bonus is always a next-dollar question.

    Defer $1,000 into a traditional 401(k) and this household’s federal income tax falls by $220. That dollar was sitting in the 22% band. Running the same move at 13.72% would understate the saving by more than a third, because the effective rate describes dollars already taxed.

    The same logic drives a retirement account decision. That is why which rate you compare is what decides Roth against traditional.

    This couple also has headroom. Taxable income of $127,800 sits $83,600 below the $211,400 floor of the 24% bracket. A raise, a bonus or a Roth conversion smaller than that stays priced at 22%.

    Advertisement

    Payroll tax sits on top and ignores brackets entirely

    Federal income tax is not the only deduction on the paycheck. Employees pay 6.2% for Social Security and 1.45% for Medicare, a combined 7.65%, and the employer pays the same again. Self-employed people pay 15.3% themselves, per the Social Security Administration.

    Those rates are flat rather than banded at this income; Medicare adds a further 0.9% on earned income above $250,000 for a couple filing jointly. The Social Security portion applies up to a wage base of $184,500 in 2026. Wages of $160,000 sit below that ceiling, so the full 7.65% lands on every dollar here. That is $12,240.

    Add it to the $17,540 of income tax and the household hands over $29,780, or 18.61% of gross wages. Even with payroll tax stacked on top, the total stays under the bracket the couple would name.

    This walk covers federal income tax, plus the employee share of payroll tax where noted. It excludes state income tax completely, which moves the result in most states. It also assumes the standard deduction rather than itemising, and it claims no credits. Every dollar here is treated as wages. Long-term capital gains and qualified dividends carry their own rate schedule and would not run through these bands.

    The short version

    Twenty-two percent touches only the last $27,000 of this household’s taxable income

    A bracket tells you where your next dollar lands. The rate you pay is a total divided by something. For this illustrative household the total is $17,540 – 13.72% of taxable income, 10.96% of gross wages. Say which denominator you used and the argument disappears.

    The walk-through runs one illustrative joint salary through the 2026 federal bands, so a household with different deductions or credits, with state or local income tax, or with income that is not all wages will land on effective rates other than the two computed here.

    Article sources

    1. IRS Rev. Proc. 2025-32 (2026 inflation-adjusted tax tables and standard deduction)
    2. Social Security and Medicare tax rates — Social Security Administration
    3. Contribution and benefit base — Social Security Administration
  • Assault Rifle vs. AR-15: Two Terms, Two Different Meanings

    Assault Rifle vs. AR-15: Two Terms, Two Different Meanings

    In this piece

    1. AR doesn’t stand for “assault rifle”
    2. Colt still owns the name
    3. Civilian versions don’t fire automatically
    4. “Assault rifle” and “assault weapon” are two different terms
    5. Why the distinction is worth knowing

    An AR-15 is a specific semi-automatic rifle Colt has sold since the 1960s — and, in everyday speech, the generic name for thousands of similar rifles built by other companies.

    Key takeaways

    • “AR” stands for ArmaLite Rifle, the company that designed it in the 1950s. It has never stood for “assault rifle.”
    • Colt holds the trademark, so almost every rifle people call an AR-15 is an AR-15-pattern rifle made by someone else.
    • Manufacturers of AR-15-pattern rifles grew from 29 in 2000 to roughly 500 by 2015.
    • Civilian versions are semi-automatic only, and machine guns made after May 19, 1986 cannot be newly registered for civilian ownership.
    • “Assault rifle” describes how a gun fires. “Assault weapon” is a statutory list of features. They are two different terms.
    A brass desk seal press standing alone on a bare wooden table in raking light, its jaws closed and its die hidden, casting a long shadow

    “AR-15” gets used two different ways in everyday conversation, and they’re not the same thing — and neither one is quite the same as “assault rifle” either. One is a specific product line with a specific history. The other is shorthand people reach for to describe almost any modern-looking semi-automatic rifle. Knowing the difference makes you harder to mislead, in either direction.

    AR doesn’t stand for “assault rifle”

    It stands for ArmaLite Rifle, named for the company that designed it in the 1950s. ArmaLite sold the design to Colt in 1959. Colt built a fully automatic version for military contracts, the M16. It also built a semi-automatic civilian version, marketed as the “AR-15.” Same lineage, two different products from the start.

    The “assault rifle” mix-up is understandable but wrong. It’s an acronym coincidence, not a description. Gun debates get heated fast. “AR” sounding like it could stand for something scarier than “ArmaLite” has kept the myth alive for decades.

    Advertisement

    Colt still owns the name — almost nobody else’s rifle is technically an “AR-15”

    Here’s the part that surprises people: Colt holds the trademark. Strictly speaking, Colt is the only company that can legally call its product an “AR-15.”

    Every other manufacturer selling a similar rifle uses its own model name instead. Bushmaster, Ruger, Smith & Wesson, and dozens more each sell a functionally comparable, semi-automatic, magazine-fed rifle built on the same general design. That list exploded once Colt’s original patents expired in the 1970s. By one count, the number of manufacturers making AR-15-pattern rifles grew from 29 in 2000 to roughly 500 by 2015.

    “AR-15” became the generic name for the whole category, the way “Kleenex” or “Xerox” did in their own product categories. Almost every rifle called an AR-15 today is really an AR-15-pattern rifle made by someone else.

    Civilian versions don’t fire automatically — and haven’t been legally new-manufactured since 1986

    This is the distinction that actually matters for understanding what these rifles do. The military M16 and M4 are select-fire: the shooter can choose semi-automatic, burst, or fully automatic. Civilian AR-15-pattern rifles are semi-automatic only — one trigger pull, one round, every time. That’s not a manufacturer choice; it’s federal law.

    Since the 1986 Hughes Amendment, machine guns made after May 19, 1986 can’t be newly registered for civilian ownership at all. A finite supply of pre-1986 registered machine guns still legally changes hands, at prices that reflect just how limited that supply is. But no civilian is walking out of a gun store today with a new, fully automatic rifle of any kind — AR-15-pattern or otherwise.

    Advertisement — advertiser disclosure

    “Assault rifle” and “assault weapon” are two different terms, not one

    This is where a lot of the confusion actually starts. They sound interchangeable. They’re not.

    Assault rifle is a military and technical term. It means a select-fire weapon, capable of semi-automatic, burst, or fully automatic fire. It’s typically chambered for an intermediate cartridge with a detachable magazine. By that definition, a semi-auto-only civilian AR-15-pattern rifle isn’t an assault rifle.

    Assault weapon is a legal and political term instead. It’s defined feature-by-feature in specific statutes, not by how the gun fires. The 1994 Federal Assault Weapons Ban classified certain semi-automatic rifles as “assault weapons” based on cosmetic and accessory features: things like a pistol grip separate from the stock, a folding or telescoping stock, a flash suppressor, or a bayonet lug. Automatic-fire capability had nothing to do with it. State assault weapon laws today generally follow that same feature-based approach, though the specific list varies by state.

    So a rifle can be legally classified as an “assault weapon” in one state, and be, mechanically, the exact same semi-automatic-only rifle sold everywhere else in the country. The classification is about listed features, not about how the trigger behaves.

    The short version

    Why the distinction is worth knowing

    None of this is an argument for or against any particular policy. It’s just accuracy. “Assault rifle,” “assault weapon,” and “AR-15” get used as if they’re one interchangeable idea. They describe three different things instead: a firing mechanism, a legal category built on cosmetic features, and a specific (mostly generic-use) product name. Getting the terms right doesn’t resolve the policy debate. But it means you’re actually debating the same thing the other person is.

    This is educational content explaining terminology and mechanics, not commentary on any pending legislation or litigation.

    Article sources

    1. Snopes — Does “AR” in AR-15 Stand for “Assault Rifle”? (acronym origin, ArmaLite/Colt history)
    2. Wikipedia — AR-15–style rifle (Colt trademark, generic-name usage, manufacturer count 2000–2015, civilian vs. military firing modes)
    3. Public Law 99-308, the Firearm Owners’ Protection Act (govinfo) — enacted May 19, 1986; § 102(9) is the Hughes Amendment, adding 18 U.S.C. 922(o), and § 110(c) makes it effective on enactment
    4. U.S. Code (Office of the Law Revision Counsel) — 18 U.S.C. § 922(o), Transfer or possession of machinegun (statute text; the effective-date note gives May 19, 1986)
    5. USCCA — Assault Rifle Definition: Cut Through the Confusion (assault rifle vs. assault weapon definitions, 1994 Federal AWB feature list)