Donald Trump on the Joe Rogan Experience #2219, fact-checked

What the evidence shows: Trump stated that across U.S. history, 92% of presidents were politicians and 8% were generals, citing Washington and Eisenhower as the only generals. Standard historical tallies do not support this figure: of the 45 individuals who have served as president, roughly 12 to 13 held general officer rank before or during their presidency (about 27 to 29%), more than three times the 8% Trump cited. Besides Washington and Eisenhower, presidents who held general rank include Andrew Jackson, William Henry Harrison, Zachary Taylor, Franklin Pierce, Andrew Johnson, Ulysses S. Grant, Rutherford B. Hayes, James Garfield, Chester Arthur, Benjamin Harrison, and William McKinley. The claim also implies a strict politician-or-general framework, but several presidents, such as Reagan, an actor and California governor, fit neither category, a point Trump himself raised about a minute later in the same conversation. The 92%/8% figures do not correspond to any standard historical count and appear to be an invented statistic used for rhetorical effect.

What the evidence shows: Trump has repeatedly claimed that China, or other foreign governments, directly pay the United States tariff revenue. In practice, tariffs are customs duties collected by U.S. Customs and Border Protection from the American companies that import the goods, not from Chinese exporters or the Chinese government. Economists and fact-checkers have found that at least part of that added cost is typically passed through to U.S. businesses and consumers in the form of higher prices, rather than being extracted from China. While tariff collections rose substantially during and after Trump's trade war with China, characterizing this revenue as money taken in from China mischaracterizes who bears the cost, since it is paid by U.S. importers at the border. FactCheck.org has documented this pattern of claims as false or misleading across multiple Trump statements, both in 2019 and again during his second term.

What the evidence shows: The Broadband Equity, Access and Deployment (BEAD) program, created by the 2021 Infrastructure Investment and Jobs Act, was funded at $42.45 billion, matching the figure Trump cited. As of his remarks and into 2025, the program had not yet connected any homes or businesses to the internet, a fact acknowledged by NTIA officials, the Trump administration itself (which cited the lack of connections as reason for a program "readjustment"), and Senate Majority Leader John Thune. The absence of connections reflects the program's multi-year statutory structure: funds were allocated to states in June 2023, initial state proposals were approved starting in April 2024, and construction was never expected to begin until 2025 or later, so zero completed connections by late 2024 was consistent with the program's planned timeline rather than solely a sign the money was wasted. A November 2025 Snopes fact-check of the same core claim (the $42.45 billion figure plus zero people connected) rated it "Mostly True." Overall, the dollar figure and zero-connections claim are accurate, but presented without the context that the delay stemmed from a required multi-year planning and approval process rather than pure waste.

What the evidence shows: Trump has repeated variations of this claim (sometimes citing $9 billion, other times $7.5 billion, and sometimes "two" instead of "three" broken chargers) about the National Electric Vehicle Infrastructure (NEVI) program created by the 2021 Infrastructure Investment and Jobs Act. Fact-checkers found the dollar figure wrong: the law allocated $5 billion for NEVI plus $2.5 billion for a separate Charging and Fueling Infrastructure grant program, totaling $7.5 billion, not $9 billion, and only a fraction of that (roughly $2.4 billion as of mid-2024) had actually been disbursed to states, not all spent on construction. The claim that only eight stations exist was also outdated: as of mid-2024 the Federal Highway Administration reported 15 operational NEVI-funded stations with 61 charging ports across eight states, and by December 2024 that had grown to 37 completed stations with 226 ports across 13 states, with hundreds more in progress. No official data supports the claim that two or three of the built stations don't work. FactCheck.org and PolitiFact both rated closely related versions of this claim false or misleading, citing the inflated dollar figure, the understated station count, and the unsupported failure-rate detail.

What the evidence shows: Trump claimed that lowering the corporate tax rate from roughly 40% to 21% under the 2017 Tax Cuts and Jobs Act (TCJA) caused revenue to rise in the first year compared to before the cut. Official federal budget data show the opposite for corporate tax receipts specifically: per the White House Office of Management and Budget's historical budget tables, corporation income tax receipts fell from $297,048 million in fiscal year 2017 (the last year at the pre-TCJA 35% statutory rate) to $204,733 million in fiscal year 2018 (the first year at the new 21% rate), a decline of about 31%. Total federal receipts did rise slightly in FY2018 versus FY2017 (from about $3,316.2 billion to about $3,329.9 billion), but that increase was driven by individual income and other tax collections growing alongside the economy, not by the corporate rate cut; corporate tax revenue itself did not increase and instead dropped sharply in the first year after the cut. The claim that the lower corporate rate produced higher revenue in its first year is contradicted by the government's own receipts data.

What the evidence shows: Trump has repeated this body-bags anecdote in multiple interviews since at least 2021, but no photograph, satellite image, or other documented evidence of body bags surrounding the Wuhan Institute of Virology has ever been produced or verified. FactCheck.org's review of the related claim that WIV lab workers fell ill in autumn 2019 found that the "body bags outside the WIV" detail is one of several unsubstantiated additions that circulate in varying, inconsistent versions of the story, with the claim traced to unnamed sources cited by a small number of former Trump administration officials rather than to any verified imagery or intelligence document. U.S. intelligence agencies remain divided on the broader question of the virus's origin: as of early 2025 the CIA, FBI, and Department of Energy assessed a lab origin as "most likely," but each with only low or moderate confidence and without agreement on which lab, while several other intelligence bodies lean toward a natural spillover origin or remain undecided. No intelligence assessment or news investigation has corroborated the specific "body bags around the Wuhan lab" detail Trump described. The broader question of COVID-19's origin remains unresolved and contested among scientists and intelligence agencies alike, but the specific body-bags-in-photographs claim remains an unverified anecdote unsupported by any documented evidence.

What the evidence shows: China has sharply accelerated coal power approvals and construction in recent years, but Trump's specific figures and framing overstate the pace. A February 2023 Global Energy Monitor/CREA report, covered by NPR, found China permitted the equivalent of about two new coal power plants per week in 2022 (106 GW total), the most in seven years, and construction continued at a high pace into 2024, with 94.5 GW of new coal projects started that year, the most in a decade. FactCheck.org, examining this same recurring Trump claim in April 2025, concluded that framing it as plants "opening" every week is misleading because an average coal plant takes roughly two years to build and come online, so permitting or construction-start pace is not the same as plants becoming operational weekly. No public dataset corroborates the specific "58 coal plants" figure Trump cites. Overall, it is directionally accurate that China is rapidly expanding coal capacity, but the precise number and the "a coal plant a week" framing are unsupported and exaggerated.

What the evidence shows: Contemporaneous reporting confirms a Chinese military incursion into Taiwan's air defense identification zone occurred around this time, but the details differ from Trump's account. According to Taiwan's defense ministry as reported by the BBC, 8 Chinese bombers (plus 4 fighter jets and 1 anti-submarine aircraft, 13 aircraft total) entered Taiwan's southwestern ADIZ on Saturday, January 23, 2021, followed by a second wave of 15 aircraft (12 fighters, 2 anti-submarine planes, 1 reconnaissance plane, no bombers) on Sunday, January 24. The combined two-day total was 28 aircraft, but only 8 of those were bombers, and the incursions occurred three to four days after Joe Biden's January 20, 2021 inauguration, not on the day Trump left office. No contemporaneous report describes 28 bombers flying over Taiwan on January 20, 2021. The claim conflates a two-day aircraft total (28, mostly fighters) with a bomber-specific count, and misstates the timing relative to Trump's exit from office.

What the evidence shows: Defense Department combat-casualty data confirms an 18-month stretch, from February 8, 2020 to August 26, 2021, in which no U.S. service member was killed in hostile action in Afghanistan, beginning shortly after the Trump administration signed the February 29, 2020 Doha agreement with Taliban negotiator Abdul Ghani Baradar. However, that stretch spans both administrations: roughly 11 months occurred under Trump and the final 7 months occurred under Biden, ending when an ISIS-K suicide bombing killed 13 U.S. service members during the August 2021 withdrawal. No 18-month span with zero combat deaths occurred entirely within Trump's presidency; fact-checkers also note four non-combat deaths and three service members wounded in action between the Doha deal and Trump leaving office. The reduction in combat deaths is generally attributed to the Doha agreement's ceasefire-style terms and phased troop drawdown rather than a single personal negotiation, and violence against Afghan government forces continued to rise sharply during this same period. Overall status: misleading, as the underlying 18-month figure is accurate but Trump's framing implies it occurred entirely under his watch and as a result of his personal deal-making alone.

What the evidence shows: The documented timeline does not support a weeks-long campaign. The U.S.-led coalition against ISIS (Operation Inherent Resolve) began in 2014 under the Obama administration, and territorial losses accrued gradually: roughly 60% of ISIS-held territory was recaptured by mid-2017, about 98% by December 2017, and 99% about a year later, per State Department figures cited by fact-checkers. The last ISIS-held territory, the village of Baghuz in eastern Syria, fell to U.S.-backed Syrian Democratic Forces on March 23, 2019 -- more than two years into Trump's presidency, not a matter of weeks. Analysts also noted that losing territory did not mean ISIS as a group was eliminated, since it retained thousands of fighters and continued attacks afterward. Framing a multi-year, multinational campaign spanning two administrations as something one person completed in weeks is not supported by the record.

What the evidence shows: The 13,099 figure traces to a September 2024 letter from ICE Deputy Director Patrick Lechleitner to Rep. Tony Gonzales, stating that many noncitizens with homicide convictions were on ICE's non-detained docket rather than in immigration custody. Fact-checkers found the data spans roughly four decades of entries into the United States, not solely the prior three years as Trump stated, and the Department of Homeland Security said the vast majority entered and had their custody status determined long before the Biden-Harris administration. Many of these individuals are not "on the loose" but are serving sentences in state, local, or federal prison custody, and the group includes legal immigrants such as green-card holders, not only people released by foreign governments emptying prisons. Court precedent, including Supreme Court rulings limiting indefinite immigration detention after a sentence is served, rather than a deliberate policy of accepting foreign prisoners, accounts for why some convicted individuals remain in the country. FactCheck.org and PolitiFact both rated versions of this claim false or misleading for conflating a decades-long cumulative ICE docket count with a claim of 13,099 murderers entering specifically over the prior three years due to countries "emptying their prisons."

What the evidence shows: Trump claimed McDonald's corporate had no record of Harris's employment and that a manager said she never worked there. Harris and her campaign have said she worked at a McDonald's on Central Avenue in Alameda, California, in the summer of 1983, running the register and fry and ice cream machines. McDonald's corporate has not confirmed or denied her employment; a company memo circulated to franchisees said McDonald's and its franchisees "don't have records for all positions dating back to the early '80s," which fact-checkers note is not equivalent to a denial. No McDonald's manager or franchise owner has gone on record stating Harris never worked there. As of publication, the claim that McDonald's or a manager confirmed Harris never worked there is unsupported by any on-the-record statement from the company.

What the evidence shows: Trump stated he built 570 miles of border wall during his presidency. Customs and Border Protection's own final tally, reported after Trump left office in January 2021, put total construction at approximately 458 miles of primary and secondary border wall system, not 570. Of that total, fact-checkers including PolitiFact found only about 52 miles were new primary barrier built where no barrier had existed before; the large majority, roughly 351 miles of primary barrier plus additional secondary barrier, replaced existing smaller or dilapidated fencing rather than extending the wall's footprint. CBP's own celebratory release used the figure "400 miles of new border wall system," a phrasing that fact-checkers say blurs the new-versus-replacement distinction. The 570-mile figure exceeds even CBP's own most generous accounting by more than 100 miles, and multiple past PolitiFact reviews of similar Trump mileage claims (450, 500 miles) rated them Half True for omitting that most construction was replacement, not new wall. Current status: misleading, both in magnitude (best documented figure is about 458 miles, not 570) and in framing (overwhelmingly replacement rather than new-terrain construction).

What the evidence shows: No independently verified government accounting supports a $200 billion national total or a $100 billion New York-specific total for migrant-related spending. For New York specifically, official city and state figures are orders of magnitude below $100 billion: the New York City Comptroller's Office reports actual migrant-related spending peaked at $3.75 billion in fiscal year 2024, with total actual-plus-budgeted spending across fiscal years 2023-2029 estimated at roughly $17.9 billion as of the June 2024 financial plan (a figure the Comptroller's Office has since revised down to about $11.8 billion in its most recent plan). Separately, the 2024 New York State budget allocated $2.4 billion in flexible migrant-support funding to the city over two fiscal years. No public dataset corroborates a $100 billion New York figure or a $200 billion national figure for migrant services.

What the evidence shows: Trump has repeated this specific claim on multiple occasions, including in an August 2025 White House statement using nearly identical language ('approximately 10 times more expensive than paper ballots'), and FactCheck.org examined that near-identical repetition directly. Matthew Weil, vice president at the Bipartisan Policy Center, told FactCheck.org the claim does not hold up: most American jurisdictions already vote on hand-marked paper ballots counted by machines that function as 'glorified scanners,' so there is no large class of all-electronic systems left to replace, and switching to watermarked paper would raise costs rather than lower them, since that paper stock is more expensive than standard ballot paper. A companion PolitiFact fact-check of the same White House remarks addressed a related claim from the same statement, that machines delay results while paper ballots would be counted the same night, and rated it False, citing election officials and researchers who said eliminating machines in favor of full hand-counting would be slower, more error-prone, and more expensive at scale, not cheaper. Neither fact-check identifies any source substantiating a precise 10x cost gap or an 8% figure for paper ballots as a share of machine costs. The claim is rated unsupported: no itemized cost data backs the specific ratio, and the most similar, directly fact-checked public repetition of this claim was found not accurate by outside experts, who described the cost dynamic as running the opposite direction from what Trump describes.

What the evidence shows: The 2005 Commission on Federal Election Reform was co-chaired by Jimmy Carter and former Secretary of State James Baker, not Senator Henry "Scoop" Jackson, who died in 1983 and could not have served on it. The commission's report stated that "absentee ballots remain the largest source of potential voter fraud" and expressed general skepticism about mail voting, but it did not conclude that mail-in ballots cannot be used; instead it called for further research on the pros and cons of voting by mail and recommended safeguards such as signature verification, citing Oregon's vote-by-mail system as having avoided significant fraud through such protections. Carter himself repeatedly stated he supported and personally used absentee/mail-in ballots, and in 2021 expressed disappointment that the report was being selectively cited to suggest it opposed mail voting. Fact-checkers, including PolitiFact and FactCheck.org, have repeatedly found similar claims by Trump that the commission's primary or sole finding was that mail-in ballots are inherently dishonest or should not be allowed to be false or a mischaracterization of the report's actual, more nuanced recommendations.

What the evidence shows: Trump's premise that China lags far behind the U.S. in nuclear weapons is consistent with Pentagon assessments: as of 2023, China's operational stockpile was estimated at over 500 warheads, versus roughly 5,000 each for the U.S. and Russia. But the Defense Department's China Military Power reports project China reaching only about 1,000 operational warheads by 2030, continuing to grow through at least 2035, still far short of the U.S. arsenal. No Pentagon or independent assessment projects China closing the gap within five years. The claim's premise is accurate, but its specific timeline is unsupported by the best available evidence.

What the evidence shows: Trump has repeatedly threatened France with 100% tariffs on wine and champagne over its digital services tax, both during his first term and again since. No available record shows the dispute was resolved by a single call in which France dropped the whole thing within minutes. After a 3% French digital tax took effect in 2019, the US and France reached a truce in January 2020 in which France agreed to delay tax collection while both sides pursued an OECD-led global tax deal, a negotiation that took roughly two weeks, not two minutes, and did not eliminate the tax. Threatened US tariffs on French goods resurfaced in mid-2020 and were formally suspended by the US Trade Representative only in January 2021, again without France repealing its tax. France's digital services tax has remained in force in the years since, having survived a legal challenge, with French lawmakers debating raising it rather than repealing it.

What the evidence shows: Trump repeated a version of this claim throughout 2024-2026, sometimes citing 90%, 95%, 98%, or "almost 100%" of the chip business as having moved to Taiwan. Industry data support a high figure only for the most advanced chip fabrication: Taiwan's TSMC produces over 90% of the world's most advanced semiconductors, largely because U.S. firms shifted decades ago to a design-only ("fabless") business model rather than because Taiwan took the business by force or deception. That narrower manufacturing statistic does not describe the whole chip industry: U.S.-headquartered companies still generate roughly half of global semiconductor revenue and dominate chip design, and overall U.S. domestic chip-manufacturing capacity, while diminished from about 37% of world output in 1990, remains meaningfully above zero and is the target of active rebuilding efforts (federal investment in Intel, incentives tied to Nvidia's China sales, new U.S. fabs). Taiwanese officials and independent industry analysts have also disputed the framing that this dominance was "stolen," attributing it instead to decades of Taiwanese capital investment and manufacturing specialization. The claim is best characterized as exaggerated: the 90%+ figure has a real basis for advanced-node manufacturing specifically, but applying it to "our business" as a whole, and describing the shift as theft, overstates and mischaracterizes what happened.