Tulsi Gabbard & Jocko Willink on the Joe Rogan Experience #1391, fact-checked

What the evidence shows: Nielsen ratings reported by Forbes show the highest-rated night of the first Democratic primary debate (NBC/MSNBC/Telemundo, June 27, 2019) drew about 18.1 million viewers, not the roughly 22 million Gabbard cites; the first night of that same two-night event drew about 15 million. The 6.6 million figure matches Nielsen numbers for the November 20, 2019 MSNBC/Washington Post debate in Atlanta, in which Gabbard participated, so that half of her comparison is accurate. Gabbard's claim correctly captures a real and steep decline in debate viewership over 2019 but overstates the peak audience by roughly 4 million viewers.

What the evidence shows: Gabbard claimed Hawaii's per-capita homelessness rate was the worst of any U.S. state. HUD's 2019 Annual Homeless Assessment Report (AHAR) to Congress, based on the January 2019 point-in-time count contemporaneous with this episode, found New York had the highest rate among states at 46 people per 10,000 residents, narrowly ahead of Hawaii at 45 per 10,000; California (38) and Oregon (38) followed. The District of Columbia, not a state, had a far higher rate of 94 per 10,000 but is excluded from state rankings. Hawaii has consistently ranked among the two or three states with the highest per-capita homelessness rates in HUD's annual counts, but it was not the single highest at the time of this claim, New York was. The claim is an exaggeration: Hawaii's homelessness crisis was severe and near the top nationally, but "worst" and "any state" overstate its position relative to New York in the data available when the episode aired.

What the evidence shows: At the time of this 2019 interview, the U.S. Census Bureau put the number of uninsured Americans at about 27.5 million in 2018, while a February 2019 Commonwealth Fund biennial survey estimated 44 million adults were underinsured (facing high out-of-pocket costs or deductibles relative to income) that same year, up from 29 million in 2010. Added together, these figures total roughly 71 to 74 million, somewhat short of Gabbard's "almost 80 million" but in the same general range given rounding and the fact that uninsured and underinsured are measured differently and are not simply additive. Both figures reflect a real and well-documented affordability problem: underinsured adults, like the uninsured, frequently skip medications, delay care, or accrue medical debt because of cost. The specific claim about Americans driving to Mexico to buy cheaper insulin reflects a widely reported phenomenon tied to high U.S. insulin list prices, though no precise national count of such trips exists. Overall, the scale of the uninsured-plus-underinsured problem is well-supported by contemporaneous government and research data, even though the exact 80-million figure is an approximation toward the high end of what the data show.

What the evidence shows: Willink attributed to the Army/Marine Corps counterinsurgency field manual (FM 3-24, published under General David Petraeus's direction in 2006) a specific claim that "the average counterinsurgency takes seven years to work itself out." FM 3-24 does discuss counterinsurgencies as typically protracted, multi-year campaigns and cautions against expecting quick victories, but the publicly available text of the manual does not contain a sourced statistic pinning the average duration at seven years; that specific figure is not traceable to the document. Independent empirical research on historical counterinsurgencies, including RAND Corporation's analysis of 30 resolved insurgencies from 1978-2008 (used partly to validate FM 3-24's tenets), finds counterinsurgency campaigns have historically run considerably longer on average, with commonly cited figures in the counterinsurgency literature clustering closer to a decade or more rather than seven years. The "seven years" figure appears to be a rounded approximation or misremembering commonly repeated in popular and military commentary rather than a verified statistic drawn directly from FM 3-24's text. Overall, the general premise that counterinsurgencies are lengthy, multi-year undertakings is well-supported, but the specific attribution of a "seven-year average" to the Petraeus-era field manual is not corroborated by the manual's actual content or by the leading empirical studies on counterinsurgency duration.

What the evidence shows: Gabbard asserted that Pete Buttigieg's 2020 presidential campaign spent $9 million on social media advertising for a roughly 4-percentage-point rise in polling. Campaign finance trackers documented that Buttigieg was among the top digital spenders in the Democratic primary field, with reported spending in the multi-million-dollar range on Facebook, Google, and other digital platforms during late 2019 and early 2020, putting the general order of magnitude in the neighborhood of the figure Gabbard cited. No reputable, allowlisted source, however, reports the specific $9 million figure or documents a 4-percentage-point polling gain directly and exclusively attributable to social media advertising; polling movement in a crowded multi-candidate primary cannot be cleanly isolated to a single spending category. Broader reporting on 2020 primary ad spending found that even much larger outlays by other candidates often produced limited or inconsistent polling returns, a dynamic consistent with the skepticism Gabbard was expressing. The precise $9 million figure and the specific causal "four percent bump" framing could not be independently verified from primary-source or reputable secondary reporting.