The art, math, and madness
of American democracy

This is PopCivics’s own anecdote, from the conversation with Ukrainian citizens that started the project, so it can’t be independently checked. Polling makes it plausible: across the 24 countries Pew surveyed in 2025, a median of 34% had confidence in Trump, and Ukrainian and Taiwanese trust in the U.S. fell sharply in 2025–26.

Nuance & counter-evidence: Global opinion is not uniform: in Pew’s 2025 survey, half or more had confidence in Trump in Nigeria (79%), Israel (69%), Kenya (64%), Hungary (53%), and India (52%). Pew didn’t survey Ukraine or Taiwan; those trends come from the Kyiv International Institute of Sociology (72% of Ukrainians called Trump’s presidency bad for Ukraine in mid-2025) and Academia Sinica (59.6% of Taiwanese disagreed that the U.S. is trustworthy in March 2025). Ukrainians still view ordinary Americans favorably (83–90%) while disapproving of U.S. leadership.

Trump-controlled hotel and casino entities filed for Chapter 11 corporate reorganization six times between 1991 and 2009—five tied to Atlantic City casinos plus the Plaza Hotel in 1992. None was a personal bankruptcy.

Nuance & counter-evidence: Chapter 11 reorganizes debt rather than liquidating a company, and Atlantic City’s wider casino market also struggled—though researchers found Trump’s properties carried unusually heavy high-yield debt (see C12). A 2014 Trump Entertainment Resorts filing occurred after he had stepped down from management.

Pew describes Trump as the first person elected U.S. president without government or military experience. Several earlier presidents never held elected office, but each had served in government or the military.

Nuance & counter-evidence: He graduated in 1964 from the New York Military Academy, a private military-style boarding school; attending it isn’t military service. During the Vietnam era he received four education deferments and, in 1968, a medical deferment. And he wasn’t wholly without political experience: he ran a months-long campaign for the Reform Party’s presidential nomination before withdrawing in February 2000, though he never held office.

On May 30, 2024, a Manhattan jury convicted Trump on 34 felony counts of falsifying business records in a single case; on Jan. 10, 2025, he received an unconditional discharge (no prison, fine, or probation). As of Oct. 14, 2026, the conviction stands while appeals proceed.

Nuance & counter-evidence: All 34 counts come from one case, about how reimbursements for a $130,000 hush-money payment were recorded. As of Oct. 14, 2026, both his direct appeal to New York’s Appellate Division, First Department, and his appeal of an Aug. 28, 2026 ruling that kept the case in state court were pending. The Second Circuit’s Nov. 2025 ruling for Trump vacated a procedural order on removal; it didn’t touch the conviction.

Per certified FEC tallies, Trump trailed in the national popular vote in 2016 (46.09% vs. Clinton’s 48.18%, −2,868,686) and 2020 (46.85% vs. Biden’s 51.31%, −7,059,526), before winning the 2024 popular vote with 77,302,580 ballots (49.80% vs. Harris’s 48.32%, +2,284,967).

Nuance & counter-evidence: Because 1.88% of 2024 ballots went to third-party and write-in candidates, Trump’s 2024 popular-vote victory was a plurality (49.80%) rather than an absolute 50%+ majority—mirroring Clinton’s 48.18% plurality in 2016.

Trump won the Electoral College 304–227 in 2016 and 312–226 in 2024, and his second term runs to January 2029.

Nuance & counter-evidence: An earlier draft of this line called him “the most consequential American president of the century.” The fact-check rated that an opinion: defensible, but premature while his term runs to 2029, and other 21st-century presidents led the country through major wars and financial crises. So the line now ends on the fact instead.

Certified FEC totals record 62,984,828 Trump votes in 2016, 74,223,975 in 2020, and 77,302,580 in 2024 (while Kamala Harris received 75,017,613 in 2024).

Nuance & counter-evidence: Keeping the election years distinct matters: 2016 foreign-interference debates occurred when Trump won 63.0 million votes, whereas his 2024 return drew 77.3 million.

Russia’s 2016 influence campaign is documented by the 2017 Intelligence Community Assessment, the Mueller report, and the bipartisan Senate Intelligence Committee. It kept going: federal prosecutors, intelligence assessments, or Treasury sanctions named Russian operations in 2018, 2020, 2022, and 2024, and in the 2026 midterms researchers tied fake videos attacking Democratic candidates to a known Russian operation. The best study of effects, Eady et al. (2023), found “no evidence of a meaningful relationship” between exposure to Russia’s Twitter posts and changes in attitudes, polarization, or vote choice.

Nuance & counter-evidence: “No one has shown” means unproven, not disproven. The Twitter study can’t rule out an effect the size of Trump’s 0.77-point Wisconsin margin in 2016, and the hacked emails and the Facebook campaign were never measured against vote choice. A 2026 study (Kim et al.) tied voter-suppression ads, from sponsors including Russia’s troll farm, to lower turnout: whether people voted, not how. The 2026 attributions come from researchers and from lawmakers describing a classified report; no public U.S. agency attribution had appeared by Oct. 26, 2026. The 2025–26 disputes over the 2017 assessment concern whether Putin wanted to help Trump, not whether Russia interfered.

This is Episode 01’s core argument, made across Acts II–V: cultural wealth-worship, commercial media incentives, Electoral College winner-take-all math, and duopoly gatekeeping. It is strongest for 2016, when the Electoral College turned a 2.87-million-vote popular-vote loss into a win.

Nuance & counter-evidence: Contingency still mattered: 38,875 switched votes across Michigan, Pennsylvania, and Wisconsin would have flipped 2016, which supports “the rules made it possible” more than “engineered.” In 2024 Trump won the popular vote (49.80%), so the Electoral College didn’t decide that outcome; the media and duopoly arguments (Acts III and V) carry his second victory.

Historians treat these as two distinct traditions: Thomas Jefferson’s 1780s yeoman farmer ideal (which distrusted concentrated wealth) and the 1830s “self-made man” popularized by Henry Clay and later critiqued by Frederick Douglass.

Nuance & counter-evidence: Neither is literally America’s oldest myth—John Winthrop’s 1630 “city upon a hill” predates both by over 150 years—and PopCivics uses “Self-Made Yeoman” deliberately to name how modern politics merges rural independence with billionaire worship.

Explicit “wealth = godliness” preaching took hold in the Gilded Age: Russell Conwell’s “Acres of Diamonds” lecture (“it is your duty to get rich”) and Bishop William Lawrence’s 1900–01 line that “Godliness is in league with riches.” In the 20th century, Norman Vincent Peale’s positive thinking and the prosperity gospel revived it, and that strand reached Trump directly through Peale, his family’s pastor, and Paula White-Cain, his longtime spiritual adviser.

Nuance & counter-evidence: It isn’t a founding-era belief: Puritan Boston fined and admonished the merchant Robert Keayne for profiteering (1639), and Jefferson contrasted a “natural aristocracy” of virtue and talents with an “artificial aristocracy founded on wealth and birth” (1813). Some of the wealth-as-merit tradition was secular, too: Andrew Carnegie defended great fortunes as the “survival of the fittest,” not divine favor. And White-Cain disputes the prosperity-gospel label, though historian Kate Bowler classifies her as a prosperity preacher.

“Pulling oneself up by one’s bootstraps” began in the 1830s as a sarcastic joke about a physical impossibility before flipping into a sincere slogan for unaided grit.

Nuance & counter-evidence: Americans are divided rather than uniformly gullible: in Pew’s 2020 survey, 65% of adults said rich people got ahead mainly through advantages (though Republicans credited hard work far more than Democrats), and Forbes classifies 71% of the 2025 Forbes 400 as self-made rather than heirs, though its label counts many head starts—making Trump’s family-seeded empire a striking counter-case.

In an 18-month investigation built on tax returns and financial records, David Barstow, Susanne Craig, and Russ Buettner documented at least $413 million (in 2018 dollars) flowing to Trump from his father’s real-estate empire—much of it through tax dodges, and including at least $60.7 million in loans. Trump had described his start as “a small loan of a million dollars.”

Nuance & counter-evidence: The figure is in 2018 dollars. The Times said “much” of the money came through tax dodges, not most; the largest single slice was his $177.3 million share of his father’s estate. Trump’s lawyer called the fraud and tax-evasion allegations false and defamatory, and no court or tax authority has publicly ruled on them.

Trump’s Atlantic City casinos went through multiple Chapter 11 restructurings after financing the $1 billion Taj Mahal with 14% junk bonds, shifting personal debt onto public shareholders and creditors who lost more than $1.5 billion while Trump continued collecting compensation and fees.

Nuance & counter-evidence: A casino’s statistical edge guarantees it wins on the bets, not that it covers its debt: four Atlantic City casinos closed in 2014 alone. So the damning benchmark is comparative. From 1997 to 2002, revenue at other Atlantic City casinos rose 18% while Trump’s fell 1% (NYT), and Temple’s Jonathan Lipson found his casinos lost 37% more employees and 33% more revenue than same-market rivals from 1997 to 2010.

Tax-record reporting showed NBC’s The Apprentice generated roughly $427 million in TV, endorsement, and licensing revenue. Separately, Trump co-founded unlicensed “Trump University” (not a mere brand license), settling state AG and class-action fraud suits for $25 million in 2016–18.

Nuance & counter-evidence: Chronology matters: The Apprentice came first and fueled deals like Trump Steaks (2007). In the Trump University litigation, New York courts ruled it operated illegally as an unlicensed educational institution, while the fraud claims settled before trial without an admission of wrongdoing.

In the American Political Science Review (2024), Eunji Kim and Shawn Patterson Jr. found the show framed Trump as a successful businessman, savvy negotiator, and tough-but-supportive mentor, and that exposure to it raised his vote share in the 2016 Republican primary. The show’s supervising editor, Jonathan Braun, told The New Yorker that Trump often fired contestants on impulse, so editors had to work backward to make the firings fit the story.

Nuance & counter-evidence: Kim & Patterson found no significant effect on the 2016 general election. A 2022 CEPR paper by Karsten Müller and Carlo Schwarz goes further, finding no effect at all, primaries included, once prior voting and NBC-viewing patterns are controlled for. A University at Buffalo survey linked heavier viewing to a one-sided “parasocial” bond with Trump, but that is a correlation. And the show relaunched an already-famous, financially battered figure rather than creating him.

In 2016, Mitt Romney called Trump “a phony, a fraud,” Marco Rubio called him “a con artist,” and Ted Cruz called him “a pathological liar.” Lindsey Graham warned that the party would “get destroyed… and we will deserve it,” and 50 former national-security officials from Republican administrations wrote that he “lacks the character, values, and experience to be President.” At the convention, Trump declared, “I alone can fix it,” and voters whose top priority was a candidate who “can bring needed change” broke 83%–14% for him.

Nuance & counter-evidence: Rubio and Cruz spoke as primary rivals, so their lines are evidence of the warnings, not proof of the charges. And not every Trump voter saw a savior: about two-thirds of 2016 voters told exit pollsters he wasn’t “honest and trustworthy,” and roughly one in five of those voted for him anyway. Among all his supporters, the most-cited reason was opposing Hillary Clinton (33%), ahead of his outsider status (27%) (Pew).

By Kagan’s estimates (via Pew), Fox News’s profit rose every year from 2006 ($216 million) to 2015 ($1.31 billion), before Trump ran. The combined profit of Fox News, CNN, and MSNBC then jumped 24% in 2016 and peaked in 2020 at $3.54 billion, the year their combined ad revenue hit $2.62 billion.

Nuance & counter-evidence: An earlier draft called it “the most lucrative outrage economy in human history,” a superlative nobody can check. The boom has also faded: the three channels’ ad revenue fell to $2.27 billion in 2022, where Pew’s series ends, and their prime-time audiences fell 34–62% from 2020 to 2024. Where that audience went next is an open question.

Adopted in 1949 and upheld in Red Lion Broadcasting (1969), the Fairness Doctrine required over-the-air broadcast licensees to cover public controversies and afford reasonable opportunity for contrasting views. Reagan vetoed codification in June 1987, and the FCC voted 4–0 on Aug. 4, 1987 (Syracuse Peace Council) to end enforcement.

Nuance & counter-evidence: The rule bound broadcast station licensees using scarce public spectrum—not print, cable, or internet outlets—and the FCC formally deleted the residual rulebook text in August 2011.

The 1987 repeal lifted one obligation from over-the-air broadcasters: seeking out and airing contrasting views on the controversial issues they covered. In 1985 the FCC had said an open, unregulated marketplace of ideas was the best way to encourage debate. Cable news grew up inside that market, where the doctrine never applied.

Nuance & counter-evidence: Whether the repeal caused the talk-radio boom is contested. Economists Thomas Hazlett and David Sosa found informational radio programming rose right after it; critics such as Free Press argue that 1990s ownership consolidation and national syndication did more to build conservative talk’s dominance. PolitiFact and Snopes both rate “the repeal created Fox News” Mostly False: Fox launched on cable in 1996, and the doctrine never reached cable channels. (A separate FCC rule did apply it to cable operators’ own local programming.)

Prime-time audiences at Fox News, CNN, and MSNBC surged in 2016 (CNN’s most-watched year ever) and peaked in 2020, when Fox drew the biggest prime-time audience in cable-news history. By 2024 the three had fallen 34%–62% from their 2020 levels.

Nuance & counter-evidence: Rather than a straight climb from 2015 to 2024, the data shows two Trump-era peaks (2016 and 2020) and a slide after. Pew’s channel-by-channel revenue series ends in 2022, and subscriber (affiliate) fees, which stayed relatively stable, cushioned the networks’ revenue.

Speaking at a Morgan Stanley investor conference on Feb. 29, 2016, Moonves said of the presidential race: “It may not be good for America, but it’s damn good for CBS.” He added that there was a lot of money in the marketplace for political advertising, and urged Trump to keep going.

Nuance & counter-evidence: Moonves was speaking specifically about paid campaign advertising revenue flowing to broadcast TV stations (not cable-news ratings), and later claimed in October 2016 that the remark was meant as a joke.

Internal documents reported in 2021 showed that from 2017, Facebook’s News Feed ranking weighted emoji reactions—“angry” included—at five times a “like.” In 2.7 million Facebook and Twitter posts, Rathje et al. (PNAS, 2021) found that posts about the opposing party were shared about twice as often, and that each word referring to the other side raised the odds of a share by 67%.

Nuance & counter-evidence: The often-repeated “500% more engagement” has no study behind it; it likely garbles Facebook’s 5× weight, a value engineers chose rather than measured, and Facebook cut reaction weights in 2020, setting “angry” to zero that September. Clicks are a different story: in 22,743 headline experiments (Robertson et al., Nature Human Behaviour, 2023), negative words raised click-through, but anger words had no significant effect and fear words reduced clicks.

That firm was mediaQuant, whose final estimate for the whole campaign was $4.96 billion in “earned media” for Trump. Earlier and narrower counts were smaller: Harvard’s Shorenstein Center valued his positive or neutral coverage in eight major outlets during 2015 at about $55 million, and called him “arguably the first bona fide media-created presidential nominee.” The New York Times’s Upshot put his early-primary free media at about $2 billion.

Nuance & counter-evidence: “Free media” is a hypothetical ad-rate valuation; no money changes hands, and methods differ (Shorenstein counted only positive and neutral coverage, and called its own figure a big underestimate). Hillary Clinton drew billions too: $3.24 billion over the whole campaign, by mediaQuant’s count.

Elbridge Gerry told the Convention that “The evils we experience flow from the excess of democracy,” George Mason compared a national popular vote to referring “a trial of colours to a blind man,” and on July 17, 1787 the Convention voted down direct popular election 9–1.

Nuance & counter-evidence: The Framers were not monolithic: James Wilson, Gouverneur Morris, Rufus King, and James Madison all argued for some form of popular choice, and the same convention made the House of Representatives directly elected.

Madison told the Convention that popular election was “the fittest in itself,” but that “the right of suffrage was much more diffusive in the Northern than the Southern States,” whose enslaved people could not vote; “the substitution of electors obviated this difficulty.” The 1790 Census confirmed the wide Northern advantage in free adult white males.

Nuance & counter-evidence: Suffrage in 1787 was set state by state, often with property requirements, and Madison was diagnosing the sectional politics on the floor rather than opposing popular election in principle. On July 25 he called it “the disproportion of qualified voters in the N. & S. States,” predicted it would shrink, and said that as a Southerner he “was willing to make the sacrifice.”

Article I, Section 2 counted free residents plus three-fifths of enslaved residents (“all other Persons”) when apportioning House seats—and Article II gave each state electors equal to its Senators plus Representatives. Under the 1792 apportionment, that formula handed slaveholding states roughly 11–14 extra House seats and presidential electors.

Nuance & counter-evidence: Precision matters: the clause did not define enslaved people morally as “three-fifths of a person”—slaveholders demanded they be counted 100% to maximize Southern seats while denying them any vote, whereas antislavery delegates argued people denied liberty should not count toward representation at all.

Small states won two flat “Senate” electors apiece and a one-state-one-vote House runoff (Sept. 4–6, 1787); slave states carried their three-fifths House bonus into the electoral count, probably decisively in 1800; and delegates kept the choice of president out of Congress’s hands.

Nuance & counter-evidence: Historians debate how much slavery was the point. Paul Finkelman and Akhil Amar argue it drove the design; Sean Wilentz calls the connection “incidental,” since the three-fifths bonus already existed if Congress chose the president; and Alexander Keyssar takes a middle position, treating slavery and large-versus-small-state power as two issues the elector system inherited from the Convention’s earlier compromises. The effect is not in dispute: even Wilentz concedes the system handed slave states the three-fifths advantage.

Forty-eight states plus D.C. use statewide winner-take-all. Only Maine (since 1972) and Nebraska (since 1992) award two electors to the statewide winner and one per congressional district; 2025–26 legislative pushes to switch Nebraska to winner-take-all failed.

Nuance & counter-evidence: Winner-take-all is state statute, not constitutional text—which is why states can lawfully change how they award electors without amending the Constitution.

By campaign events (National Popular Vote’s tracker), Wesleyan Media Project ad tracking, and ABC News/538’s analysis, those seven states (93 electoral votes) were the 2024 battleground; Trump carried all seven by margins ranging from 0.86 points in Wisconsin to 5.53 points in Arizona.

Nuance & counter-evidence: Maine’s 2nd and Nebraska’s 2nd congressional districts also drew targeted competition, and the swing-state map shifts across eras (Florida and Ohio hosted 119 combined events in 2016 vs. zero in 2024).

Per July 1, 2024 U.S. Census population estimates, 81.7% of Americans lived outside the seven swing states—while 94% of 2024 general-election candidate/VP campaign events (246 of 262) took place inside those seven states.

Nuance & counter-evidence: Safe-state voters aren’t literally irrelevant—their predictable majorities supply the first ~218 Republican and ~225 Democratic electoral votes, and national media reaches every state—but campaigns rarely visit or tailor field operations to them.

In 2024, 6,081,697 Californians voted for Trump and 4,835,250 Texans voted for Harris (more Harris votes than any state but California, New York included). Every ballot was counted and certified, yet under statewide winner-take-all each group moved zero electoral votes.

Nuance & counter-evidence: Those ballots are never “thrown out”—they are legally certified and counted in the national popular total—which is why saying they “elect zero electors” is far more accurate than saying they are “discarded.”

The combined winning margin across the decisive trio of states was 77,744 votes in 2016 (PA, WI, MI), 42,918 in 2020 (WI, GA, AZ—for a 269–269 tie), and 229,766 in 2024 (WI, MI, PA)—meaning ~39,000, ~21,500, or ~115,000 vote-switchers would have altered the Electoral College outcome.

Nuance & counter-evidence: There’s no evidence the deciding voters were “undecided” or concentrated in suburban counties. Pew’s validated-voter study found 2024’s shift came mostly from who turned out, not from voters switching sides, and Trump’s largest 2024 gains came in more urban, more racially diverse places (ABC/538).

Official FEC 2016 totals: Clinton 65,853,514 (48.18%) vs. Trump 62,984,828 (46.09%)—a popular-vote margin of 2,868,686 votes (2.10 percentage points), while Trump won the Electoral College 304–227 (306–232 pledged).

Nuance & counter-evidence: Neither candidate topped 50% in 2016 because third-party and write-in candidates drew 5.73% of the national vote.

Certified 2016 state margins: Pennsylvania +44,292, Wisconsin +22,748 (the tipping-point state), and Michigan +10,704—totaling 77,744 votes. Had 38,875 voters across those three states switched from Trump to Clinton, Clinton would have carried all three and the presidency.

Nuance & counter-evidence: Narrow state margins also favored Clinton in New Hampshire (+2,736) and Nevada (+27,202), illustrating how winner-take-all magnifies razor-thin pluralities in whichever states sit near the 270 line.

The Constitution fixes how many electors each state gets and lets each legislature decide how to appoint them. Statewide winner-take-all (the “general ticket”) isn’t in it: Pennsylvania used it in the first election, and states adopted it one by one, mostly between 1800 and 1832, as McPherson v. Blacker (1892) recounts. The 12th Amendment (1804) changed the balloting—separate votes for president and vice president—not the apportionment.

Nuance & counter-evidence: Jefferson himself backed Virginia’s 1800 switch to a general ticket so Virginia wouldn’t split its clout while rival states voted as blocs—the exact prisoners’ dilemma that the National Popular Vote Compact seeks to unwind via state law.

In 2024, Trump won both the Electoral College (312–226) and the national popular vote, 77,302,580 to 75,017,613 (49.80% to 48.32%)—a plurality, not a majority. The Electoral College’s Republican tilt, about 2.9 points in 2016 and 3.8 in 2020, shrank to about 0.2 points in 2024 as Trump gained ground in big non-swing states like New York, New Jersey, and California.

Nuance & counter-evidence: On the votes actually cast, a National Popular Vote Compact would have elected Trump in 2024 too; campaigns would have run differently under it, so nobody can know the totals in that world. The structural point survives: the Electoral College can elect the popular-vote loser, as it did in 2016, and 2020 was a near miss—about 21,000 switched votes in three states would have produced a 269–269 tie.

In Gallup’s December 2023 reading, a record-low 28% of Americans were satisfied with the way democracy is working, down from 61% in 1984. That shows legitimacy eroding across the board, though not what caused it. And the DNC itself argued in court that its charter’s promise of “impartiality and evenhandedness” was a mere political promise (see C36).

Nuance & counter-evidence: Imperfect primary rules and attempting to overturn a certified election are not morally equivalent—the point is strategic credibility with disillusioned voters, and gatekeeping habits appear in both major parties (see C37).

In Wilding v. DNC Services Corp., the DNC argued that its charter’s promise of “impartiality and evenhandedness” was a mere political promise, and its lawyer said the party could choose its nominee in “back rooms like they used to.” The district court dismissed the suit in 2017 for lack of standing; the 11th Circuit affirmed in 2019, finding that the donor plaintiffs had standing but hadn’t pleaded fraud with enough particularity. In 2024, after Biden, the presumptive nominee, withdrew on July 21, delegates elected in his primaries nominated Harris by virtual roll call; no voter had cast a 2024 primary ballot for her.

Nuance & counter-evidence: Since the early-1970s reforms, the pledged-delegate leader has won every Democratic nomination—including Hillary Clinton in 2016, by more than 3.7 million primary votes—and in 2018 Democrats barred superdelegates from the first ballot. The 2024 roll call was open on paper: any candidate with 300 delegate signatures could have entered, and only Harris qualified.

Ahead of 2024, Democrats held zero primary debates and Florida’s Democratic Party certified only Joe Biden—canceling the state’s presidential primary. Ahead of 2020, the RNC pledged its undivided backing to Trump in Jan. 2019, and state GOPs in South Carolina, Nevada, Kansas, and Arizona scrapped or skipped their primaries and caucuses.

Nuance & counter-evidence: No incumbent president of either party has ever joined a primary debate, and open-seat primaries (such as Democrats in 2016 and 2020, or Republicans in 2016) remain fiercely contested.

Independent and minor-party presidential bids face 50 separate state petition regimes (e.g., 219,403 valid signatures in California for 2024), aggressive ballot-line lawsuits (such as knocking the Green Party off Nevada’s 2024 ballot), and the bipartisan Commission on Presidential Debates’ 15% national polling threshold.

Nuance & counter-evidence: Steep ballot laws come from legislatures run by both parties: Democratic-run California’s and Republican-run Florida’s, Texas’s, and North Carolina’s. A few states ask for very few signatures (800 in New Jersey; 7,500 in Georgia, set by a court). And in 2024 the Biden campaign refused CPD debates altogether.

Pew’s validated-voter study found voters without a four-year college degree backed Trump 56%–42% in 2024, double his 2016 margin among them. Hispanic voters went 51%–48% for Harris, down from 61%–36% for Biden in 2020, and Trump’s share of Black voters rose from 8% to 15%.

Nuance & counter-evidence: Whether Democrats “took the working class for granted” is interpretation. Sen. Bernie Sanders blamed “a Democratic Party which has abandoned working class people” after the 2024 loss; defenders point to Biden’s pro-labor record, including the first picket-line visit by a sitting president, and argue that inflation, culture, and immigration drove the shift.

In California Democratic Party v. Jones (2000), the Supreme Court reaffirmed parties’ First Amendment right as private associations to define membership—yet states print, staff, and pay for party primaries and grant the two major parties automatic general-election ballot lines.

Nuance & counter-evidence: Because state-administered primaries entangle party rules with public ballots, the Supreme Court held in Smith v. Allwright (1944) that parties cannot use “private club” status to enforce racial exclusions in primaries.

Citizens United (Jan. 2010) freed corporate and union independent spending, and SpeechNow.org v. FEC (Mar. 2010) created super PACs funded by unlimited contributions. In the 2024 cycle, the top 100 individual donors gave $2.4 billion, about 16% of all federal election spending (OpenSecrets data via Mother Jones). Crypto companies put more than $119 million into federal races, mostly through the Fairshake super PAC, which backed candidates in both parties.

Nuance & counter-evidence: Individual megadonors usually pick a side: 2024’s biggest donor, Elon Musk, gave more than $291 million over the cycle, all of it, per OpenSecrets, to Republican candidates, PACs, and other outside groups. Super PACs must report their donors, though money routed through nonprofits that don’t disclose theirs can hide the source. And voters across parties say big donors have outsize influence, but no study we found shows that perceived donor capture drove Trump votes.

Lesson 2 of On Tyranny (“Defend institutions”) says institutions fall one after another unless each is defended from the start, and tells readers to pick one—a court, a newspaper, a law, a labor union—and take its side. Lesson 3 (“Beware the one-party state”) urges readers to support a multiple-party system and defend the rules of democratic elections.

Nuance & counter-evidence: Attribution clarity: “Institutions do not protect themselves” is Snyder’s line; the corollary that you can’t defend a machine you don’t understand is PopCivics’s own addition. Snyder contrasts a multiple-party system with one-party rule; that isn’t an endorsement of RCV or the compact, which work through lawful state-level channels.

Three 1787 clauses protected slaveholding (Three-Fifths apportionment, the Article V-shielded 1808 slave-trade clause, and the Fugitive Slave Clause), and the original Constitution contained no affirmative right to vote—deferring suffrage to states until the 15th, 19th, 24th, and 26th Amendments barred specific exclusions.

Nuance & counter-evidence: “City upon a hill” comes from Puritan governor John Winthrop’s 1630 sermon (“A Model of Christian Charity”) as a warning against failure, 157 years before Philadelphia; Ronald Reagan popularized the optimistic “shining city upon a hill.” The minority-rule bug is concrete: the Electoral College winner received fewer popular votes than an opponent in 1876, 1888, 2000, and 2016 (National Archives).

Alaska uses top-four RCV statewide, Maine uses RCV for federal races and state primaries, D.C. implemented RCV in June 2026, and dozens of cities use ranked ballots. Empirical analysis of 182 multi-round U.S. RCV elections (Graham-Squire & McCune, 2023) found spoiler and Condorcet anomalies in under 2% of races.

Nuance & counter-evidence: Honest limits: 19 states now ban RCV (with Indiana and Ohio joining in early 2026), all four pro-RCV state ballot measures lost in Nov. 2024, and Alaska votes on repeal (Ballot Measure 2) on Nov. 3, 2026. Moreover, single-winner RCV alone doesn’t end two-party dominance—Australia’s House has ranked ballots since 1918 and two major blocs still won 137 of 150 seats in 2025; reformers such as Lee Drutman argue that multiparty representation takes proportional, multi-member districts.

“America 2.0” is PopCivics’s own editorial shorthand—not the official legal title (“Agreement Among the States to Elect the President by National Popular Vote”) or National Popular Vote Inc.’s slogan.

Nuance & counter-evidence: Technically the compact is a state-law software patch running on top of the 1787 Electoral College rather than a brand-new constitution: electors still meet and vote, and Article IV of the compact terminates the agreement if the Electoral College is ever abolished by amendment.

Virginia Gov. Abigail Spanberger signed HB 965 / SB 322 on April 13, 2026, bringing the compact to 18 states plus D.C. and 222 of the 270 electoral votes needed (48 short). Using states’ Article II elector-appointment power, members pledge their electors to whichever ticket wins the largest national popular-vote total once members hold 270+ EVs on July 20 of an election year.

Nuance & counter-evidence: Three essential caveats: (1) it awards electors to the national plurality winner (even under 50%); (2) whether Article I’s Compact Clause requires congressional consent is legally unresolved (CRS R43823), and legal challenges are likely once it reaches 270; and (3) Brennan Center 2030 Census projections estimate current members could lose ~7 EVs starting in 2032.

The U.S. Census Bureau’s Vintage 2025 estimate places the resident population at 341.8 million (up from 331.4 million in 2020), while 41 of 100 senators can deny the 60 votes needed to end debate on most bills.

Nuance & counter-evidence: Who uses the 41-vote block flips whenever Senate control changes: since Jan. 2025, the 47-seat Democratic minority represents states home to roughly 53% of Americans.

The filibuster is an internal chamber rule resting on Article I, Section 5 (“Each House may determine the Rules of its Proceedings”), made possible when the Senate dropped its “previous question” motion in 1806 (by mistake, in political scientist Sarah Binder’s account). Rule XXII, adopted in 1917, created cloture: the way to end one.

Nuance & counter-evidence: “Not in the Constitution” does not make it unconstitutional: federal courts have dismissed challenges on standing and rulemaking grounds, leaving reform in the hands of a Senate majority.

From the 1890s through 1964, Southern senators repeatedly wielded talking filibusters to kill anti-lynching, anti-poll-tax, and civil-rights bills—accounting for roughly half of all measures defeated by filibuster between 1917 and 1994.

Nuance & counter-evidence: Civil rights was its deadliest historical target rather than its origin (1806), and modern everyday filibustering across nearly all domestic policy surged after the 1970s two-track calendar and 1975 switch from two-thirds voting to 60 seated senators.

Since the 1970s two-track system, a contested bill is simply set aside unless supporters muster 60 votes for cloture. Individual “holds” relayed via letter or cloakroom hotline block unanimous consent and cost scarce floor days, while 41 no/absent votes on cloture stop passage.

Nuance & counter-evidence: A lone senator’s email or letter cannot permanently veto a bill that has 60 votes behind it—it only forces the majority leader to burn floor time on the cloture clock.

Using 2025 Census estimates, the 20 smallest states plus half of Utah’s delegation equal 10.7% of the U.S. population—the theoretical floor for a blocking minority under equal state suffrage and a 60-vote rule.

Nuance & counter-evidence: In practice, those smallest states split across both parties (VT, DE, RI, and HI alongside WY, ND, SD, and AK, with Maine’s delegation split), majorities carved out nominations via precedent in 2013, 2017, and Sept. 2025, and budget reconciliation bills bypass the 60-vote hurdle.

Because Rule XXII counts against all seated senators rather than those present and voting, absences function like “no” votes on cloture. Executive and judicial nominations (including Supreme Court justices) require only a simple majority, while formally amending Senate rules under Rule XXII requires two-thirds voting.

Nuance & counter-evidence: When cloture fails on the initial “motion to proceed,” the Senate never even opens formal floor debate on the bill itself.

Despite White House calls in 2025–26 to scrap the legislative filibuster over shutdown and election bills, Senate Republicans kept the 60-vote threshold for legislation intact while using the simple-majority precedent in Sept. 2025 to confirm groups of executive nominees en bloc.

Nuance & counter-evidence: Roughly three-quarters of all 119th-Congress cloture roll calls through Oct. 2026 involved nominations (where 51 votes suffice); legislative obstruction appears in the 44 failed legislative cloture votes over that span.