Political blind spots

Wallace, Goldman, Polyakova (NBER, 2022) analyzed county-level death certificates and found age-adjusted excess mortality was 43% higher in counties that voted Trump than in counties that voted Biden, concentrated after COVID vaccines became widely available. The gap was near-zero before vaccines rolled out.

Reuters/Ipsos polling (April 2024) finds 62% of Republicans believe Biden's 2020 win was illegitimate, despite 60+ court rulings finding no evidence of fraud, including from Trump-appointed judges. Ten state recounts (including GOP-led ones in Georgia and Arizona) confirmed original totals.

Pew Research 2024: 78% of Democrats and Dem-leaners say climate change is a major threat to the country. Among Republicans: 24%. The gap is larger than on any other Pew-tracked issue and has widened since 2015. Scientific consensus across every major national academy is unambiguous that warming is human-caused.

Pew's long-running confidence-in-institutions series shows Republican confidence in scientists fell from 43% saying "great deal" in 2019 to 23% in 2023. Democratic confidence held roughly flat at 65-70%. The split predates COVID but accelerated during it.

CDC WONDER age-adjusted firearm death rates (2022) averaged 19.4 per 100,000 in the 10 states Trump won by >10 points vs 8.7 per 100,000 in the 10 states Biden won by >10 points. Includes suicides (~55% of the total), homicides, and accidents. Gap correlates with state gun-law strictness.

PRRI 2024 "American Values Atlas": 34% of white evangelical Protestants and 27% of Republicans overall agree "America should be a Christian nation." Only 14% of Americans overall agree. The belief predicts support for restrictions on non-Christian public expression and opposition to church-state separation.

Internal platform studies leaked via the Meta Papers (2021), Twitter Files (2022-23), and the Bloomberg "Gemini cheat code" reporting (2024): outrage-coded right-wing content consistently generates 4-6× the engagement of equivalent left-wing content. The asymmetry is why AI-generated political content defaults toward conservative framing when optimizing for reach.

Pew Research 2023 + Gallup Generations data: among white women 18-29, 56% of liberals report a diagnosed mental-health condition vs 27% of conservatives in the same age/race/gender bracket. Jonathan Haidt and Jean Twenge attribute the gap to social-media exposure plus ideology-coded framings of adversity as trauma. The pattern is robust across multiple surveys.

Cato / YouGov 2022 survey: Democrats as a group estimate median US household income at roughly $35,000 and estimate median wealth at $30,000. Actual figures from Census and Fed SCF are $75K and $192K respectively. Republicans over-estimate both in the opposite direction. Neither party is well-calibrated on the distribution their policy views rest on.

Mitchell Langbert / Heterodox Academy review of 51 top liberal-arts colleges (2023): across history, English, sociology, and anthropology departments, Democrat-to-Republican faculty ratios average 15:1 to 70:1. The monoculture is documented to narrow research questions and bias peer review (Duarte et al., Behavioral and Brain Sciences, 2015).

HUD 2024 Annual Homeless Assessment Report: the 10 cities with the highest per-capita unsheltered homelessness (LA, SF, Seattle, Portland, NYC, Oakland, San Diego, San Jose, DC, Boston) all have Democratic mayors, most for 20+ years. Correlation is real; causation is contested. Climate, housing policy, cost of living, and state involuntary-commitment rules all confound. But blue-city governance has not solved the problem on its own timeline.

Pew Research 2021 and Collaborative Multi-Racial Post-Election Survey 2020: 31% of white college-educated liberals supported reducing police-department budgets in 2020; only 11% of Black voters in the communities most affected by policing did. Rob Henderson labels this the "luxury belief" pattern, positions whose cost is borne mostly by people who don't hold them.

Cato / YouGov National Survey (2020, repeated 2023): in 2020, 62% of strong liberals said "the political climate prevents me from sharing my views;" by 2023 it was 66%. Strong conservatives rose 53%→54%. Self-censorship grew 2.8× faster on the left over the interval. The most common cause cited: career retaliation in Dem-leaning industries.

The "trust the experts" posture has notable failures: the 2020-22 suppression of lab-leak discussion (DOE + FBI later concluded lab leak plausible); NIH gain-of-function funding opacity; the UK NHS Cass Review walking back pediatric gender medicine that US progressive institutions had adopted without equivalent evidence review; WHO's initial masking guidance; Iraq WMD (bipartisan but Dem-coded in hindsight). Pattern: institutional consensus hardened before evidence settled, and dissenters were labeled conspiracy theorists.

In 2016, Jill Stein's vote totals exceeded Trump's winning margin in each of Michigan (Stein 51,463, margin 10,704), Wisconsin (Stein 31,072, margin 22,748), and Pennsylvania (Stein 49,941, margin 44,292). The three states together decided the Electoral College. Counterfactual vote re-distribution is unknowable, but the first-past-the-post math is: in a two-party-dominant system without ranked-choice voting, a Green Party run with >0.5% support in a swing state is structurally a Republican assist. Ralph Nader's 97,488 Florida votes in 2000 (Bush margin: 537) is the canonical earlier case.

The Green Party US platform calls for phasing out nuclear energy. IPCC AR6 WG3 (2022) pathways that meet the 1.5°C warming goal assume roughly a doubling of global nuclear capacity by 2050 in the median scenario; almost no scenario achieves the goal with nuclear shrinking. Germany's 2011-2023 nuclear phase-out coincided with a decade of higher emissions per kWh than if the plants had stayed online (peer-reviewed: Jarvis, Deschenes, Jha, AER 2022). Hansen, Caldeira, Emanuel, Wigley (2013 open letter) argued nuclear is essential to deep decarbonization. The anti-nuclear position is the clearest case where the party's policy and the party's own stated climate goal are in measurable tension.

The Green Party US platform opposes genetically modified crops as hazardous to human health. 275+ independent scientific organizations (National Academies of Sciences 2016, WHO, EU Joint Research Centre, AAAS, AMA) have concluded that currently-marketed GMO foods pose no greater risk than conventionally-bred counterparts. The epistemic structure (distrust of corporate-linked science, privileging precautionary intuition over peer-reviewed consensus) mirrors the pattern this page documents on the right regarding climate change. The evidence-consensus gap is comparable; the polarity is reversed.

Senate Intelligence Committee Report Vol 5 (2020, bipartisan) documented overlap between Jill Stein's 2016 campaign messaging and Russian information operations, including social-amplification of Stein content by the Internet Research Agency. Stein attended the December 2015 RT 10th-anniversary gala in Moscow and was seated at President Putin's table. Stein voluntarily testified to the Committee in 2019. The Committee did not find criminal wrongdoing; it did find the messaging alignment was operationally useful to a hostile foreign service. Stein's 2016 campaign also raised over $7M for post-election recounts in MI/WI/PA that, if successful, would have reopened the certification process Russia was actively contesting.

The Green Party of the United States has never held a US House or Senate seat. Highest-ever presidential vote share: 2.74% (Nader, 2000). Currently ~150 local officeholders nationwide; zero governorships, zero federal seats. Has appeared on all 50 state presidential ballots only twice (2000, 2024, with 2024 subject to significant Democratic Party ballot-access litigation). The infrastructure gap means the party functions structurally as a protest-vote vehicle in presidential races rather than as a building coalition for down-ballot representation, even where ranked-choice voting (Maine, Alaska) would favor the latter.

Green Party platform endorses steady-state or degrowth economics, arguing GDP growth is incompatible with emissions reduction. Our World in Data + World Bank: 35+ countries achieved absolute decoupling (GDP up, production + consumption emissions down) over 2000-2022, including the UK (−42% CO2 / +35% GDP), Germany, Denmark, Sweden. Green growth is therefore empirically possible rather than theoretically impossible. Most IPCC pathways to 1.5°C assume continued global-South GDP growth with emissions falling via efficiency + deployment, not shrinkage. The degrowth position is a policy preference; the empirical claim that it is the only viable path is not supported by the recent data.

Gary Johnson received 3.28% of the 2016 popular vote, the highest Libertarian Party share in a presidential election. In Michigan, Johnson received 172,136 votes (Trump's margin: 10,704); in Pennsylvania 146,715 (margin 44,292); in Wisconsin 106,674 (margin 22,748). Post-election polling on second-choice preferences is mixed and Johnson voters' ideological lean was not uniformly Republican, but as with the Green Party, the raw vote-transfer math materially exceeded the deciding margin in multiple states. In a first-past-the-post system, a Libertarian run at >2% share is structurally a Democratic assist in closely-contested elections.

ANES 2016 + Cato Libertarian-voter surveys: Libertarian-identified voters are approximately 85% white and 75% male, making the Libertarian Party the narrowest demographic profile of any significant US political coalition. The Republican coalition is ~70% white; the Democratic coalition is ~58% white. The demographic skew is a structural ceiling problem: the coalition has never broken 4% nationally in a presidential race, and the under-representation of women and non-white voters in the base contradicts the party's self-description as the ideology of universal individual liberty.

At the May 2022 national convention, the Mises Caucus won a majority of the Libertarian Party National Committee. Subsequent platform changes removed or weakened language on abortion rights, immigration, and LGBTQ protections that had been in the party platform since the 1970s. Founding-era figures (Bill Weld, Gary Johnson, Jo Jorgensen) publicly distanced themselves; several state parties (New Mexico, Virginia, Massachusetts) disaffiliated. The internal contradiction is documented by the party's own platform-history archive: the party built around personal liberty moved toward positions that restrict it, driven by an intra-party faction fight that the original coalition lost.

Libertarian-aligned economists and advocates (Greenspan, Phil Gramm, the Cato regulatory-critique program) championed the Commodity Futures Modernization Act (2000) that exempted credit-default swaps and most OTC derivatives from oversight, the 2004 SEC leverage-rule change, and hands-off housing-market regulation through the 2000s. The 2008 financial crisis is widely attributed in the peer-reviewed literature (FCIC Report 2011, Gorton, Admati/Hellwig) to the interaction of those regulatory gaps with housing-GSE dynamics and Fed policy. Greenspan himself, testifying before Congress in October 2008, acknowledged his ideology-based framework "had a flaw." The experiment ran; the results are on the record.

The Commonwealth Fund 2024 comparison of 11 high-income countries ranked the US last overall on healthcare system performance, despite spending ~17% of GDP (roughly double the peer average). The US system is also the most market-based of the 11, highest private-insurance share, lowest government-payer share, most price-negotiation decentralization. The empirical test for "more markets → better health outcomes" is running, and the most market-based peer system is the worst-performing. Libertarian proposals for further decentralization / HSA-based reform have not been implemented at the scale needed to test whether the direction reverses the result.

Office of Rail and Road data: UK rail real fares (inflation-adjusted) rose approximately 60% between the 1995-97 privatization and the 2020 pre-pandemic baseline, while reliability metrics declined and public subsidy to operators grew. By 2021-24 the UK government had brought multiple franchises back under public operation (LNER, Northern, TransPennine, Southeastern). Similar pattern: Chilean water privatization (1989-2019) produced some of the world's highest water tariffs and was a named grievance in the 2019-20 unrest. The systematic empirical case that privatization improves outcomes across sectors is not supported; the case on specific assets varies widely.

2024 outside-spending by Super PACs hit a record ~$4.5 billion, roughly 7× the first post-Citizens United cycle (2012: ~$600M) and rivaling what the candidates' own committees raise. Citizens United v. FEC and SpeechNow v. FEC (both 2010) created the legal architecture that makes unlimited independent expenditures by corporations, unions, and nonprofits constitutionally protected.

Elon Musk gave approximately $277M to Trump-aligned groups during the 2024 cycle, primarily through America PAC. It is the largest documented individual donation in a single US election cycle. Timothy Mellon gave ~$172M to MAGA Inc. Kenneth Griffin and the Adelson estate gave in the $75-$100M range each. Ten donors accounted for more than a third of all GOP-aligned Super PAC money.

Michael Bloomberg gave approximately $92M to Democratic-aligned groups in 2024. The Pritzker family, Reid Hoffman (LinkedIn cofounder, ~$26M), and George Soros-aligned committees rounded out the top Democratic mega-donors. The Democratic donor class is smaller in per-capita gift size but similarly concentrated; both parties now depend structurally on eight-figure individual gifts.

The top 100 mega-donors of the 2024 cycle gave more combined than the bottom ~3.7 million small-donor contributions (under $200 each) put together. One hundred individuals have more financial influence on the American political system than several million voters giving at whatever level they can afford. The distribution is structurally Pareto, not Gaussian.

501(c)(4) 'social welfare' nonprofits can donate unlimited amounts to Super PACs without disclosing their donors. These flows are known as dark money because the original source is legally concealed. The 2024 cycle set a record at ~$1.1 billion; Democratic-aligned dark money (~$570M) slightly exceeded Republican-aligned (~$460M) for the first time in the post-Citizens United era. Dark-money growth has outpaced disclosed PAC spending.

OpenSecrets analysis of 2024 US House races found the candidate with more combined campaign + outside spending won roughly 91% of races. The relationship held in both directions and across both parties. Money does not guarantee victory, but the correlation is strong enough that the funding gap determines outcomes in the large majority of competitive seats. The rational response for incumbents, challengers, and parties alike is to chase larger checks.

September 2024 DOJ indictment: RT (Russian state media) funneled approximately $10 million to Tenet Media, a Tennessee-registered LLC, which in turn paid six US right-wing influencers (Tim Pool, Dave Rubin, Benny Johnson, Lauren Southern, Matt Christiansen, Tayler Hansen) to produce content. Russian operatives Kostiantyn Kalashnikov and Elena Afanasyeva concealed the source behind a fictional 'Eduard Grigoriann' European-wealth persona. The influencers were not charged; they claimed no knowledge of the Russian source. Tenet dissolved within a week of the indictment.

Per the DOJ indictment, the videos Tenet Media produced, 'most of which support the goals of the Russian government,' accumulated 16 million views on YouTube before the scheme was disrupted. AG Merrick Garland said the operation's goal, as part of the broader 'Doppelganger' Russian campaign, was 'amplifying US domestic divisions in order to weaken US opposition to core Russian interests, particularly its ongoing war in Ukraine.' The Biden administration simultaneously sanctioned 10 individuals and entities and seized 32 internet domains used by the operation.

Way to Win, a major Democratic donor group, spent more than $9.1 million on social-media influencers during the 2024 presidential election. The money was not listed in the FEC disclosure portals that normally show political spending, because the FEC has ruled that campaign payments to creators for political messaging are not 'public communications' requiring disclosure. Parallel outlays on the Republican side ran through less-identifiable intermediaries and are harder to count precisely, but the practice is bipartisan.

FEC filings show the Harris campaign paid approximately $4 million to Village Marketing Agency, a firm that connects clients with social-media influencers. A separate firm, People First, paid individual influencers ranging from $200 to $100,000 per political post. The Harris campaign also paid Oprah's Harpo Productions $1M for 'event production', Beyoncé's company Parkwood ~$165K, and Springsteen's Thrill Hill ~$75K. None of the creators' social posts were required to say 'paid for by'.

The 2024 FEC advisory found that paid posts by social-media creators about political candidates or issues do not qualify as 'public communications' under federal campaign law, so no 'paid for by' disclosure is required. That gap applies symmetrically: foreign ops, Republican PACs, Democratic PACs, and presidential campaigns all exploit it. Viewers have no way to distinguish paid from organic political content from a given creator. A bipartisan proposal to close the gap has been introduced every Congress since 2018 and has never reached a floor vote.

Part of the Biden administration's September 2024 counter-operation. Seized domains mimicked legitimate Western outlets (Washington Post, Der Spiegel, Le Monde) but pushed Russian narratives. The Office of the Director of National Intelligence separately assessed that Russia, China, and Iran all ran US-targeted influence operations in 2024, with Russia accounting for the largest share. Successor networks have continued operating through 2025.

On April 21, 2026, Virginia voters approved a constitutional amendment (51.5% to 48.6%) bypassing the state's bipartisan redistricting commission. Democrats are expected to gain four US House seats. The vote was the Democratic response to Trump's push for GOP-led states to redraw maps outside the decennial census cycle. Running scoreboard: Democrats +10 (California 5, Virginia 4, Utah 1 court-ordered), Republicans +5 (Texas). DeSantis called a Florida special session for April 28; Ohio and Missouri are reportedly next. The 2025-26 push broke a ~40-year informal rule that states redistrict only after a census.

The 2025-26 mid-decade war began in summer 2025 when Texas Republicans, at Trump's urging, passed a new congressional map projected to shift five seats to the GOP. California Democrats answered with Proposition 50 in November 2025: a ballot measure that bypassed the state's independent redistricting commission to approve a Democratic-drawn map, also projected to shift five seats. The Supreme Court upheld California's map in February 2026 and declined to block Texas's. Both maps will be in effect for the 2026 midterms. The direct partisan result cancels out; the precedent, that both parties will abandon independent commissions when it suits them, is permanent.

Cook Political Report's Partisan Voting Index: roughly 90% of the 435 US House seats are rated R+5 or D+5 or stronger, meaning the general-election outcome is effectively predetermined. Only 25-40 seats are rated toss-up in any given cycle. The combined effect of geographic self-sorting, decennial gerrymandering, and incumbent-protection drawing produces a House where the vast majority of members cannot lose to the opposite party.

OpenSecrets: 96% of US House incumbents who sought re-election in 2024 won. The rate has exceeded 90% in every cycle since 1964 and 85% every cycle since 1946. Combined with safe districts and low primary turnout, the US House is closer to tenured employment than to a competitive labor market. A moderate challenger cannot unseat an extreme incumbent through general-election mechanics alone; they have to win a primary with a small, ideological electorate.

Bipartisan Policy Center: average 2022 congressional primary turnout was approximately 17% of eligible voters, vs 60-65% for general elections. Primary voters are measurably more ideological than general voters (Pew 2023, American National Election Studies). Combined with ~90% safe districts, the actual selectors of most US House members are ~10% of the electorate, skewed to the ideological poles of each party. The rational candidate response is to run toward the base, not the center.

Gilens & Page (Perspectives on Politics, 2014) analyzed 1,779 policy decisions 1981-2002 against the preferences of median voters, economic elites, and organized interest groups. They found a near-zero statistical effect of median-voter preference on policy outcomes once elite and interest-group preferences were controlled for. The paper has been replicated, contested, and re-replicated across multiple datasets; the core finding survives. It is the empirical center of the 'American oligarchy thesis'.

On April 22, 2026, Kalshi suspended and fined three US congressional candidates for 'political insider trading', wagering on the prediction-market contracts for their own campaigns. Mark Moran (VA independent Senate) was fined over $6,200 and refused a negotiated settlement. Matt Klein (MN-02 Democratic primary) paid ~$540. Ezekiel Enriquez (TX-21 Republican primary) placed bets under $100. All three were suspended for 5 years. Kalshi + Polymarket together now run hundreds of millions of dollars in annualized political event-contract volume; Congress is actively scrutinizing the structure. It is the first prominent enforcement action in the category and the clearest signal yet that the prediction-market architecture imports the insider-trading problem it was pitched against.

Public Citizen and the Center for Responsive Politics: approximately 50% of retiring US senators and ~40% of retiring House members register as lobbyists or take lobby-adjacent executive roles within five years of leaving office. Average compensation rises 4-5×. The product for sale is the knowledge, access, and personal relationships they accumulated while in office. Both parties are represented in the pipeline at roughly equal rates. Combined with the Gilens & Page finding above, the policy-output machinery is: mega-donor funds campaign → candidate wins safe seat → legislator votes with elite/interest-group preferences → legislator retires into a lobby firm paid by those same groups.

Seven swing states (MI, PA, WI, GA, AZ, NV, NC) decided the 2024 presidential election. Roughly 96% of 2024 general-election campaign events, and a similar share of advertising, were concentrated in those seven. The other 43 states received essentially no campaign attention. If you live in California, New York, Illinois, Texas, Florida, or most of the rest of the country, no presidential candidate spent any resources trying to persuade you, because winner-take-all Electoral College allocation makes your state's outcome a foregone conclusion weeks before election day.

A Wyoming voter's Electoral College influence is approximately 3.7× that of a California voter, because each state gets two Senate-based electors regardless of population. Small states are systematically overweighted. Compounding this: 48 of 50 states use winner-take-all allocation of electors (only Maine and Nebraska split by district). A Democrat's vote in Wyoming and a Republican's vote in California literally do not contribute to the Electoral College outcome; they are counted in the popular-vote total only. The 2000 and 2016 elections both produced popular-vote losers who won the presidency. The National Popular Vote Interstate Compact has been enacted by 17 states + DC totaling 209 of the 270 EC votes needed for it to take effect.

As of March 2026 the Supreme Court had issued 35 emergency-docket ('shadow docket') orders in cases related to the second Trump administration; the administration won ~80% of them, usually without oral argument and with little or no written explanation. Justice Kagan, dissenting, wrote the docket has been used this year 'to transfer government authority from Congress to the President, and thus to reshape the Nation's separation of powers.' Historical shadow-docket use averaged ~2 rulings per term; under the Roberts Court on Trump cases it is running an order of magnitude higher.

Pew Research (September 2025): 48% of Americans have a favorable opinion of the Supreme Court, down from 70% five years earlier. 81% of Republicans approve; 19% of Democrats do. 56% of Americans say the court's decisions are based mainly on politics rather than law (Marquette Law School, consistent reading since July 2023). 43% say the court is 'too conservative,' the highest in Gallup's trend. A binding code of ethics is supported by 86% of Democrats and 62% of Republicans.